feat: 架构修复 - P0/P1 问题全面修复

P0 修复: - runtimeStore: 添加缺失的 lastDayHistory 字段 - Gateway/RuntimeService: 状态同步改为内存优先，消除 glob 竞态 - App.jsx: 从 3075 行重构到 ~500 行，提取 8 个独立文件 P1 修复: - CORS: 4 个服务改为从环境变量读取允许 origins - MarketStore: 改为模块级单例模式 - Domain 层: 删除 trading thin wrapper，保留 news 真实逻辑 - 测试: 补齐 77 个 gateway/runtime 测试新增文件: - backend/tests/test_gateway.py (43 tests) - frontend/src/hooks/useWebSocketHandler.js - frontend/src/hooks/useStockRequestCallbacks.js - frontend/src/hooks/useAgentCallbacks.js - frontend/src/hooks/useRuntimeCallbacks.js - frontend/src/hooks/useWatchlistCallbacks.js - frontend/src/components/TickerBar.jsx - frontend/src/components/HeaderRight.jsx - frontend/src/components/ChartTabs.jsx Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
fix(frontend): 添加缺失的 lastDayHistory 字段到 runtimeStore
2026-03-23 18:45:57 +08:00 · 2026-03-23 18:20:29 +08:00 · 2026-03-23 18:12:32 +08:00 · 2026-03-23 18:03:23 +08:00 · 2026-03-23 17:46:06 +08:00 · 2026-03-23 17:45:39 +08:00
442 changed files with 44428 additions and 3050 deletions
--- a/.omc/project-memory.json
+++ b/.omc/project-memory.json
@@ -1,7 +1,7 @@
 {
  "version": "1.0.0",
-  "lastScanned": 1773304964541,
-  "projectRoot": "/Users/cillin/workspeace/agentscope-samples/evotraders",
+  "lastScanned": 1773938154948,
+  "projectRoot": "/Users/cillin/workspeace/evotraders",
  "techStack": {
    "languages": [
      {
@@ -11,6 +11,14 @@
        "markers": [
          "pyproject.toml"
        ]
+      },
+      {
+        "name": "C/C++",
+        "version": null,
+        "confidence": "high",
+        "markers": [
+          "Makefile"
+        ]
      }
    ],
    "frameworks": [
@@ -24,8 +32,8 @@
    "runtime": null
  },
  "build": {
-    "buildCommand": null,
-    "testCommand": "pytest",
+    "buildCommand": "make build",
+    "testCommand": "make test",
    "lintCommand": "ruff check",
    "devCommand": null,
    "scripts": {}
@@ -40,7 +48,8 @@
    "isMonorepo": false,
    "workspaces": [],
    "mainDirectories": [
-      "docs"
+      "docs",
+      "scripts"
    ],
    "gitBranches": {
      "defaultBranch": "main",
@@ -49,29 +58,64 @@
  },
  "customNotes": [],
  "directoryMap": {
+    "agent-service": {
+      "path": "agent-service",
+      "purpose": null,
+      "fileCount": 2,
+      "lastAccessed": 1773938154941,
+      "keyFiles": [
+        "Dockerfile",
+        "requirements.txt"
+      ]
+    },
    "backend": {
      "path": "backend",
      "purpose": null,
-      "fileCount": 3,
-      "lastAccessed": 1773304964533,
+      "fileCount": 5,
+      "lastAccessed": 1773938154941,
      "keyFiles": [
        "__init__.py",
+        "app.py",
        "cli.py",
+        "gateway_server.py",
        "main.py"
      ]
    },
+    "backtest": {
+      "path": "backtest",
+      "purpose": null,
+      "fileCount": 0,
+      "lastAccessed": 1773938154941,
+      "keyFiles": []
+    },
+    "data": {
+      "path": "data",
+      "purpose": "Data files",
+      "fileCount": 1,
+      "lastAccessed": 1773938154941,
+      "keyFiles": [
+        "market_research.db"
+      ]
+    },
+    "deploy": {
+      "path": "deploy",
+      "purpose": null,
+      "fileCount": 0,
+      "lastAccessed": 1773938154942,
+      "keyFiles": []
+    },
    "docs": {
      "path": "docs",
      "purpose": "Documentation",
      "fileCount": 0,
-      "lastAccessed": 1773304964533,
+      "lastAccessed": 1773938154942,
      "keyFiles": []
    },
    "evotraders.egg-info": {
      "path": "evotraders.egg-info",
      "purpose": null,
      "fileCount": 6,
-      "lastAccessed": 1773304964534,
+      "lastAccessed": 1773938154942,
      "keyFiles": [
        "PKG-INFO",
        "SOURCES.txt",
@@ -83,8 +127,8 @@
    "frontend": {
      "path": "frontend",
      "purpose": null,
-      "fileCount": 12,
-      "lastAccessed": 1773304964535,
+      "fileCount": 13,
+      "lastAccessed": 1773938154942,
      "keyFiles": [
        "README.md",
        "components.json",
@@ -93,239 +137,488 @@
        "index.css"
      ]
    },
+    "live": {
+      "path": "live",
+      "purpose": null,
+      "fileCount": 0,
+      "lastAccessed": 1773938154943,
+      "keyFiles": []
+    },
+    "logs": {
+      "path": "logs",
+      "purpose": null,
+      "fileCount": 7,
+      "lastAccessed": 1773938154943,
+      "keyFiles": [
+        "2026-03-16_00-48-03.log",
+        "2026-03-18_23-17-29.log",
+        "2026-03-18_23-17-30.2026-03-18_23-17-30_000801.log.zip",
+        "2026-03-18_23-17-30.log",
+        "2026-03-19_00-18-04.log"
+      ]
+    },
+    "news-service": {
+      "path": "news-service",
+      "purpose": null,
+      "fileCount": 3,
+      "lastAccessed": 1773938154943,
+      "keyFiles": [
+        "Dockerfile",
+        "requirements.txt"
+      ]
+    },
+    "reference": {
+      "path": "reference",
+      "purpose": null,
+      "fileCount": 0,
+      "lastAccessed": 1773938154943,
+      "keyFiles": []
+    },
+    "runs": {
+      "path": "runs",
+      "purpose": null,
+      "fileCount": 0,
+      "lastAccessed": 1773938154944,
+      "keyFiles": []
+    },
+    "scripts": {
+      "path": "scripts",
+      "purpose": "Build/utility scripts",
+      "fileCount": 1,
+      "lastAccessed": 1773938154944,
+      "keyFiles": [
+        "run_prod.sh"
+      ]
+    },
+    "services": {
+      "path": "services",
+      "purpose": "Business logic services",
+      "fileCount": 1,
+      "lastAccessed": 1773938154944,
+      "keyFiles": [
+        "README.md"
+      ]
+    },
+    "shared": {
+      "path": "shared",
+      "purpose": null,
+      "fileCount": 0,
+      "lastAccessed": 1773938154944,
+      "keyFiles": []
+    },
+    "trading-service": {
+      "path": "trading-service",
+      "purpose": null,
+      "fileCount": 4,
+      "lastAccessed": 1773938154944,
+      "keyFiles": [
+        "Dockerfile",
+        "README.md",
+        "requirements.txt"
+      ]
+    },
+    "workspaces": {
+      "path": "workspaces",
+      "purpose": null,
+      "fileCount": 0,
+      "lastAccessed": 1773938154944,
+      "keyFiles": []
+    },
+    "agent-service/src": {
+      "path": "agent-service/src",
+      "purpose": "Source code",
+      "fileCount": 5,
+      "lastAccessed": 1773938154944,
+      "keyFiles": [
+        "__init__.py",
+        "config.py",
+        "main.py"
+      ]
+    },
+    "backend/api": {
+      "path": "backend/api",
+      "purpose": "API routes",
+      "fileCount": 5,
+      "lastAccessed": 1773938154944,
+      "keyFiles": [
+        "__init__.py",
+        "agents.py",
+        "guard.py"
+      ]
+    },
    "backend/config": {
      "path": "backend/config",
      "purpose": "Configuration files",
-      "fileCount": 4,
-      "lastAccessed": 1773304964535,
+      "fileCount": 6,
+      "lastAccessed": 1773938154944,
      "keyFiles": [
        "__init__.py",
-        "constants.py",
-        "data_config.py"
+        "agent_profiles.yaml",
+        "bootstrap_config.py"
      ]
    },
    "backend/data": {
      "path": "backend/data",
      "purpose": "Data files",
-      "fileCount": 7,
-      "lastAccessed": 1773304964536,
+      "fileCount": 13,
+      "lastAccessed": 1773938154944,
      "keyFiles": [
        "__init__.py",
        "cache.py",
        "historical_price_manager.py"
      ]
    },
-    "backend/services": {
-      "path": "backend/services",
-      "purpose": "Business logic services",
-      "fileCount": 4,
-      "lastAccessed": 1773304964536,
-      "keyFiles": [
-        "__init__.py",
-        "gateway.py",
-        "market.py"
-      ]
-    },
-    "backend/tests": {
-      "path": "backend/tests",
-      "purpose": "Test files",
-      "fileCount": 4,
-      "lastAccessed": 1773304964536,
-      "keyFiles": [
-        "__init__.py",
-        "test_agents.py",
-        "test_market_service.py"
-      ]
-    },
    "docs/assets": {
      "path": "docs/assets",
      "purpose": "Static assets",
      "fileCount": 5,
-      "lastAccessed": 1773304964536,
+      "lastAccessed": 1773938154944,
      "keyFiles": [
        "dashboard.jpg",
        "evotraders_demo.gif",
        "evotraders_logo.jpg"
      ]
    },
-    "frontend/public": {
-      "path": "frontend/public",
-      "purpose": "Public files",
-      "fileCount": 1,
-      "lastAccessed": 1773304964538,
+    "frontend/dist": {
+      "path": "frontend/dist",
+      "purpose": "Distribution/build output",
+      "fileCount": 2,
+      "lastAccessed": 1773938154945,
      "keyFiles": [
+        "index.html",
        "trading_logo.png"
      ]
+    },
+    "frontend/node_modules": {
+      "path": "frontend/node_modules",
+      "purpose": "Dependencies",
+      "fileCount": 1,
+      "lastAccessed": 1773938154947,
+      "keyFiles": []
+    },
+    "news-service/src": {
+      "path": "news-service/src",
+      "purpose": "Source code",
+      "fileCount": 3,
+      "lastAccessed": 1773938154948,
+      "keyFiles": [
+        "__init__.py",
+        "config.py",
+        "main.py"
+      ]
+    },
+    "trading-service/src": {
+      "path": "trading-service/src",
+      "purpose": "Source code",
+      "fileCount": 8,
+      "lastAccessed": 1773938154948,
+      "keyFiles": [
+        "__init__.py",
+        "config.py",
+        "main.py"
+      ]
    }
  },
  "hotPaths": [
    {
-      "path": "frontend/src/components/StatisticsView.jsx",
-      "accessCount": 22,
-      "lastAccessed": 1773310044545,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/components/AgentCard.jsx",
+      "path": "backend/agents/factory.py",
      "accessCount": 17,
-      "lastAccessed": 1773309995177,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/App.jsx",
-      "accessCount": 12,
-      "lastAccessed": 1773309849392,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/components/AgentFeed.jsx",
-      "accessCount": 12,
-      "lastAccessed": 1773309960022,
-      "type": "file"
-    },
-    {
-      "path": ".env",
-      "accessCount": 7,
-      "lastAccessed": 1773308950505,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/components/RoomView.jsx",
-      "accessCount": 7,
-      "lastAccessed": 1773309864236,
-      "type": "file"
-    },
-    {
-      "path": "backend/tools/analysis_tools.py",
-      "accessCount": 5,
-      "lastAccessed": 1773312271446,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/components/Header.jsx",
-      "accessCount": 4,
-      "lastAccessed": 1773309827069,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/components/AboutModal.jsx",
-      "accessCount": 4,
-      "lastAccessed": 1773310093371,
-      "type": "file"
-    },
-    {
-      "path": "backend/agents/prompts/analyst/personas.yaml",
-      "accessCount": 4,
-      "lastAccessed": 1773312049213,
-      "type": "file"
-    },
-    {
-      "path": "backend/agents/prompts/analyst/system.md",
-      "accessCount": 4,
-      "lastAccessed": 1773312049696,
-      "type": "file"
-    },
-    {
-      "path": "backend/agents/prompts/portfolio_manager/system.md",
-      "accessCount": 4,
-      "lastAccessed": 1773312050326,
-      "type": "file"
-    },
-    {
-      "path": "backend/agents/prompts/risk_manager/system.md",
-      "accessCount": 4,
-      "lastAccessed": 1773312050782,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/config/constants.js",
-      "accessCount": 3,
-      "lastAccessed": 1773309824671,
-      "type": "file"
-    },
-    {
-      "path": "frontend/src/components/RulesView.jsx",
-      "accessCount": 3,
-      "lastAccessed": 1773310061939,
+      "lastAccessed": 1773939950376,
      "type": "file"
    },
    {
      "path": "backend",
-      "accessCount": 3,
-      "lastAccessed": 1773312200721,
+      "accessCount": 16,
+      "lastAccessed": 1773940042371,
      "type": "directory"
    },
+    {
+      "path": "",
+      "accessCount": 13,
+      "lastAccessed": 1773939899611,
+      "type": "directory"
+    },
+    {
+      "path": "backend/main.py",
+      "accessCount": 7,
+      "lastAccessed": 1773939993951,
+      "type": "file"
+    },
+    {
+      "path": "backend/gateway_server.py",
+      "accessCount": 7,
+      "lastAccessed": 1773940004402,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/main.py",
+      "accessCount": 5,
+      "lastAccessed": 1773938385662,
+      "type": "file"
+    },
+    {
+      "path": "backend/core/pipeline.py",
+      "accessCount": 5,
+      "lastAccessed": 1773940024933,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/enrich/news_enricher.py",
+      "accessCount": 4,
+      "lastAccessed": 1773938508417,
+      "type": "file"
+    },
+    {
+      "path": "start-dev.sh",
+      "accessCount": 4,
+      "lastAccessed": 1773939259381,
+      "type": "file"
+    },
+    {
+      "path": "services/README.md",
+      "accessCount": 4,
+      "lastAccessed": 1773939281935,
+      "type": "file"
+    },
+    {
+      "path": "backend/app.py",
+      "accessCount": 4,
+      "lastAccessed": 1773939648215,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/routes/news.py",
+      "accessCount": 3,
+      "lastAccessed": 1773938438928,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news",
+      "accessCount": 3,
+      "lastAccessed": 1773938468730,
+      "type": "directory"
+    },
+    {
+      "path": "frontend/src/config/constants.js",
+      "accessCount": 3,
+      "lastAccessed": 1773939204395,
+      "type": "file"
+    },
    {
      "path": "backend/services/gateway.py",
+      "accessCount": 3,
+      "lastAccessed": 1773939672930,
+      "type": "file"
+    },
+    {
+      "path": "backend/core/__init__.py",
+      "accessCount": 3,
+      "lastAccessed": 1773939963627,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/trading/main.py",
      "accessCount": 2,
-      "lastAccessed": 1773312232905,
+      "lastAccessed": 1773938360736,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/agents/main.py",
+      "accessCount": 2,
+      "lastAccessed": 1773938361040,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/trading/data/__init__.py",
+      "accessCount": 2,
+      "lastAccessed": 1773938402496,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/explain/__init__.py",
+      "accessCount": 2,
+      "lastAccessed": 1773938460019,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/enrich/__init__.py",
+      "accessCount": 2,
+      "lastAccessed": 1773938465216,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/explain/range_explainer.py",
+      "accessCount": 2,
+      "lastAccessed": 1773938481152,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/enrich/llm_enricher.py",
+      "accessCount": 2,
+      "lastAccessed": 1773938499885,
+      "type": "file"
+    },
+    {
+      "path": "CLAUDE.md",
+      "accessCount": 2,
+      "lastAccessed": 1773939273598,
+      "type": "file"
+    },
+    {
+      "path": "backend/agents/__init__.py",
+      "accessCount": 2,
+      "lastAccessed": 1773939883015,
+      "type": "file"
+    },
+    {
+      "path": "backend/agents/agent_core.py",
+      "accessCount": 2,
+      "lastAccessed": 1773939886997,
+      "type": "file"
+    },
+    {
+      "path": "Makefile",
+      "accessCount": 1,
+      "lastAccessed": 1773938226307,
+      "type": "file"
+    },
+    {
+      "path": "docker-compose.yml",
+      "accessCount": 1,
+      "lastAccessed": 1773938226360,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/news/shared/trading_client.py",
+      "accessCount": 1,
+      "lastAccessed": 1773938370618,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/agents",
+      "accessCount": 1,
+      "lastAccessed": 1773938397772,
      "type": "directory"
    },
    {
-      "path": "README.md",
+      "path": "backend/services/trading",
      "accessCount": 1,
-      "lastAccessed": 1773305013217,
+      "lastAccessed": 1773938397823,
+      "type": "directory"
+    },
+    {
+      "path": "backend/services",
+      "accessCount": 1,
+      "lastAccessed": 1773938405541,
+      "type": "directory"
+    },
+    {
+      "path": "backend/services/news/config.py",
+      "accessCount": 1,
+      "lastAccessed": 1773938638664,
      "type": "file"
    },
    {
-      "path": "README_zh.md",
+      "path": "shared/client/news_client.py",
      "accessCount": 1,
-      "lastAccessed": 1773305013274,
+      "lastAccessed": 1773938638715,
      "type": "file"
    },
    {
-      "path": "env.template",
+      "path": "shared/client/trading_client.py",
      "accessCount": 1,
-      "lastAccessed": 1773305019965,
+      "lastAccessed": 1773938638770,
+      "type": "file"
+    },
+    {
+      "path": "backend/api",
+      "accessCount": 1,
+      "lastAccessed": 1773938669143,
+      "type": "directory"
+    },
+    {
+      "path": "frontend",
+      "accessCount": 1,
+      "lastAccessed": 1773938669195,
+      "type": "directory"
+    },
+    {
+      "path": ".env.example",
+      "accessCount": 1,
+      "lastAccessed": 1773938849397,
      "type": "file"
    },
    {
      "path": "frontend/src/services/websocket.js",
      "accessCount": 1,
-      "lastAccessed": 1773309324302,
+      "lastAccessed": 1773938849448,
      "type": "file"
    },
    {
-      "path": "backend/config/data_config.py",
+      "path": "frontend/src/services/runtimeApi.js",
      "accessCount": 1,
-      "lastAccessed": 1773309324414,
+      "lastAccessed": 1773938849500,
      "type": "file"
    },
    {
-      "path": "backend/cli.py",
+      "path": "backend/services/agents/routes/websocket.py",
      "accessCount": 1,
-      "lastAccessed": 1773309336899,
-      "type": "directory"
-    },
-    {
-      "path": "backend/agents/portfolio_manager.py",
-      "accessCount": 1,
-      "lastAccessed": 1773311956562,
+      "lastAccessed": 1773939001692,
      "type": "file"
    },
    {
-      "path": "backend/agents/risk_manager.py",
+      "path": "backend/services/agents/routes/agents.py",
      "accessCount": 1,
-      "lastAccessed": 1773311956760,
+      "lastAccessed": 1773939016291,
+      "type": "file"
+    },
+    {
+      "path": "backend/services/agents/routes/run.py",
+      "accessCount": 1,
+      "lastAccessed": 1773939016343,
+      "type": "file"
+    },
+    {
+      "path": "backend/__init__.py",
+      "accessCount": 1,
+      "lastAccessed": 1773939648323,
+      "type": "file"
+    },
+    {
+      "path": "backend/api/__init__.py",
+      "accessCount": 1,
+      "lastAccessed": 1773939658650,
+      "type": "file"
+    },
+    {
+      "path": "backend/runtime/__init__.py",
+      "accessCount": 1,
+      "lastAccessed": 1773939658687,
+      "type": "file"
+    },
+    {
+      "path": "backend/agents/base/evo_agent.py",
+      "accessCount": 1,
+      "lastAccessed": 1773939664916,
      "type": "file"
    },
    {
      "path": "backend/agents/analyst.py",
      "accessCount": 1,
-      "lastAccessed": 1773311963222,
+      "lastAccessed": 1773939664967,
      "type": "file"
    },
    {
-      "path": "backend/tools",
+      "path": "backend/agents/base/hooks.py",
      "accessCount": 1,
-      "lastAccessed": 1773312289643,
-      "type": "directory"
+      "lastAccessed": 1773939672727,
+      "type": "file"
    },
    {
-      "path": "backend/tools/data_tools.py",
+      "path": "pyproject.toml",
      "accessCount": 1,
-      "lastAccessed": 1773312293851,
-      "type": "directory"
+      "lastAccessed": 1773939672778,
+      "type": "file"
    }
  ],
  "userDirectives": []
--- a/.omc/state/hud-state.json
+++ b/.omc/state/hud-state.json
@@ -1,6 +1,6 @@
 {
-  "timestamp": "2026-03-12T20:33:59.497Z",
+  "timestamp": "2026-03-19T16:36:52.471Z",
  "backgroundTasks": [],
-  "sessionStartTimestamp": "2026-03-12T14:19:33.615Z",
-  "sessionId": "73b0d597-0141-4873-9d0e-2b60e4e0635e"
+  "sessionStartTimestamp": "2026-03-19T16:36:42.224Z",
+  "sessionId": "ef02339a-1eec-4c7a-95ac-c8cfa0b5067d"
 }
--- a/.omc/state/hud-stdin-cache.json
+++ b/.omc/state/hud-stdin-cache.json
@@ -1 +1 @@
-{"session_id":"73b0d597-0141-4873-9d0e-2b60e4e0635e","transcript_path":"/Users/cillin/.claude/projects/-Users-cillin-workspeace-agentscope-samples-evotraders/73b0d597-0141-4873-9d0e-2b60e4e0635e.jsonl","cwd":"/Users/cillin/workspeace/agentscope-samples/evotraders","model":{"id":"kimi-for-coding","display_name":"kimi-for-coding"},"workspace":{"current_dir":"/Users/cillin/workspeace/agentscope-samples/evotraders","project_dir":"/Users/cillin/workspeace/agentscope-samples/evotraders","added_dirs":["/Users/cillin/workspeace/agentscope-samples/EvoTraders","/Users/cillin/workspeace/agentscope-samples/evotraders"]},"version":"2.1.63","output_style":{"name":"default"},"cost":{"total_cost_usd":6.822239999999999,"total_duration_ms":42679588,"total_api_duration_ms":1223637,"total_lines_added":275,"total_lines_removed":240},"context_window":{"total_input_tokens":654274,"total_output_tokens":27014,"context_window_size":200000,"current_usage":{"input_tokens":48465,"output_tokens":0,"cache_creation_input_tokens":0,"cache_read_input_tokens":0},"used_percentage":24,"remaining_percentage":76},"exceeds_200k_tokens":false}
+{"session_id":"ef02339a-1eec-4c7a-95ac-c8cfa0b5067d","transcript_path":"/Users/cillin/.claude/projects/-Users-cillin-workspeace-evotraders/ef02339a-1eec-4c7a-95ac-c8cfa0b5067d.jsonl","cwd":"/Users/cillin/workspeace/evotraders","model":{"id":"MiniMax-M2.7-highspeed","display_name":"MiniMax-M2.7-highspeed"},"workspace":{"current_dir":"/Users/cillin/workspeace/evotraders","project_dir":"/Users/cillin/workspeace/evotraders","added_dirs":[]},"version":"2.1.78","output_style":{"name":"default"},"cost":{"total_cost_usd":17.458779250000003,"total_duration_ms":1866224,"total_api_duration_ms":1188013,"total_lines_added":257,"total_lines_removed":290},"context_window":{"total_input_tokens":195204,"total_output_tokens":48917,"context_window_size":200000,"current_usage":{"input_tokens":481,"output_tokens":0,"cache_creation_input_tokens":149,"cache_read_input_tokens":163286},"used_percentage":82,"remaining_percentage":18},"exceeds_200k_tokens":false}
--- a/.omc/state/idle-notif-cooldown.json
+++ b/.omc/state/idle-notif-cooldown.json
@@ -1,3 +1,3 @@
 {
-  "lastSentAt": "2026-03-12T20:31:37.362Z"
+  "lastSentAt": "2026-03-19T17:02:32.170Z"
 }
--- a/.omc/state/subagent-tracking.json
+++ b/.omc/state/subagent-tracking.json
@@ -1,26 +1,17 @@
 {
  "agents": [
    {
-      "agent_id": "a4090d26a45ac828d",
-      "agent_type": "oh-my-claudecode:executor",
-      "started_at": "2026-03-12T10:02:38.238Z",
+      "agent_id": "a8305a91e192b2196",
+      "agent_type": "Explore",
+      "started_at": "2026-03-19T17:00:33.284Z",
      "parent_mode": "none",
      "status": "completed",
-      "completed_at": "2026-03-12T10:10:59.192Z",
-      "duration_ms": 500954
-    },
-    {
-      "agent_id": "af87583ef76a4df30",
-      "agent_type": "oh-my-claudecode:executor",
-      "started_at": "2026-03-12T10:40:04.409Z",
-      "parent_mode": "none",
-      "status": "completed",
-      "completed_at": "2026-03-12T10:41:17.387Z",
-      "duration_ms": 72978
+      "completed_at": "2026-03-19T17:02:19.439Z",
+      "duration_ms": 106155
    }
  ],
-  "total_spawned": 2,
-  "total_completed": 2,
+  "total_spawned": 1,
+  "total_completed": 1,
  "total_failed": 0,
-  "last_updated": "2026-03-12T10:41:17.490Z"
+  "last_updated": "2026-03-19T17:02:39.175Z"
 }
--- a/CLAUDE.md
+++ b/CLAUDE.md
@@ -0,0 +1,302 @@
+# CLAUDE.md
+
+本文件为 Claude Code (claude.ai/code) 在此代码库中工作时提供指导。
+
+## 项目概述
+
+EvoTraders 是一个自进化多智能体交易系统，由 6 个 AI Agent（4 名分析师 + 投资经理 + 风控经理）协作完成交易决策。Agent 基于 AgentScope 框架构建，配合 ReMe 记忆系统实现持续学习。
+
+## 常用命令
+
+### Backend (Python)
+
+```bash
+# 安装依赖
+uv pip install -e .
+
+# 运行命令
+evotraders backtest --start 2025-11-01 --end 2025-12-01  # 回测模式
+evotraders backtest --start 2025-11-01 --end 2025-12-01 --enable-memory  # 带记忆回测
+evotraders live                              # 实盘交易
+evotraders live --mock                       # 模拟/测试模式
+evotraders live -t 22:30                     # 定时每日交易
+evotraders frontend                          # 启动可视化界面
+
+# 开发服务器
+./start-dev.sh  # 启动全部 4 个微服务
+
+# 单独启动某个服务
+python -m uvicorn backend.apps.agent_service:app --host 0.0.0.0 --port 8000 --reload
+python -m uvicorn backend.apps.runtime_service:app --host 0.0.0.0 --port 8003 --reload
+python -m uvicorn backend.apps.trading_service:app --host 0.0.0.0 --port 8001 --reload
+python -m uvicorn backend.apps.news_service:app --host 0.0.0.0 --port 8002 --reload
+
+# 测试
+pytest backend/tests                          # 运行全部测试
+pytest backend/tests/test_news_service_app.py -v  # 运行单个测试文件
+pytest backend/tests/test_news_service_app.py::test_news_service_routes_are_exposed -v  # 运行单个测试
+```
+
+### Frontend (React)
+
+```bash
+cd frontend
+npm run dev          # Vite 开发服务器 (http://localhost:5173)
+npm run build        # 生产构建
+npm run lint         # ESLint 检查
+npm run lint:fix     # ESLint 自动修复
+npm run test         # Vitest 单元测试
+npm run test:watch   # 监听模式
+```
+
+## 架构概览
+
+### 微服务架构 (`backend/apps/`)
+
+项目采用 split-first 微服务架构，4 个独立的 FastAPI 服务：
+
+| 服务 | 入口 | 端口 | 职责 |
+|------|------|------|------|
+| agent_service | `backend.apps.agent_service:app` | 8000 | Agent 生命周期、工作区管理 |
+| runtime_service | `backend.apps.runtime_service:app` | 8003 | 运行时配置、任务启动 |
+| trading_service | `backend.apps.trading_service:app` | 8001 | 市场数据、交易操作 |
+| news_service | `backend.apps.news_service:app` | 8002 | 新闻、新闻富化、解释功能 |
+
+服务间通过环境变量通信（详见 `start-dev.sh`）：
+```bash
+export TRADING_SERVICE_URL=http://localhost:8001
+export NEWS_SERVICE_URL=http://localhost:8002
+export RUNTIME_SERVICE_URL=http://localhost:8003
+```
+
+### Gateway 网关 (`backend/services/gateway.py`)
+
+Gateway 是统一的请求路由器，根据路径前缀将请求转发到对应的微服务：
+- `/control/*` → agent_service
+- `/runtime/*` → runtime_service
+- `/trading/*` → trading_service
+- `/news/*` → news_service
+
+新增接口时应注册到对应的 service app，而非直接添加到 gateway。
+
+### 共享客户端 (`shared/client/`)
+
+统一的服务客户端库，所有前端和后端服务间通信都使用此处定义的客户端：
+
+| 客户端 | 用途 |
+|--------|------|
+| `ControlPlaneClient` | Agent 服务通信 |
+| `RuntimeServiceClient` | 运行时服务通信 |
+| `TradingServiceClient` | 交易服务通信 |
+| `NewsServiceClient` | 新闻服务通信 |
+
+### 领域层 (`backend/domains/`)
+
+业务逻辑按领域分离：
+
+- `news.py` - 新闻领域操作
+- `trading.py` - 交易领域操作
+
+## 后端结构
+
+```
+backend/
+├── agents/           # 多智能体实现
+│   ├── base/         # 核心类、Hooks、评估
+│   │   ├── evo_agent.py        # 基于 AgentScope 的核心实现
+│   │   ├── hooks.py             # 生命周期 Hooks
+│   │   │   ├── BootstrapHook           # 启动初始化
+│   │   │   ├── MemoryCompactionHook    # 内存压缩（基于 CoPaw）
+│   │   │   ├── HeartbeatHook           # 心跳检测
+│   │   │   └── WorkspaceWatchHook      # 工作区监控
+│   │   ├── evaluation_hook.py   # 执行后评估
+│   │   ├── skill_adaptation_hook.py  # 动态技能适配
+│   │   └── tool_guard.py        # 工具调用守卫
+│   ├── prompts/      # Agent 提示词和角色定义
+│   │   ├── analyst/personas.yaml  # 分析师角色配置
+│   │   └── portfolio_manager/
+│   ├── team/         # 团队协作逻辑
+│   │   ├── registry.py        # Agent 注册表
+│   │   ├── coordinator.py     # 协作协调器
+│   │   ├── messenger.py       # 消息传递
+│   │   └── task_delegator.py  # 任务分发
+│   ├── factory.py     # Agent 实例工厂
+│   ├── skills_manager.py  # 技能加载管理（6 种作用域）
+│   └── toolkit_factory.py  # 工具集工厂
+├── apps/             # 微服务入口（split-first）
+│   ├── agent_service.py
+│   ├── runtime_service.py
+│   ├── trading_service.py
+│   └── news_service.py
+├── domains/          # 领域业务逻辑
+│   ├── news.py
+│   └── trading.py
+├── services/         # Gateway 和辅助服务
+│   ├── gateway.py           # 统一路由网关
+│   ├── gateway_*.py         # Gateway 子模块
+│   └── market.py            # 市场数据服务
+├── api/              # FastAPI 端点
+├── config/           # 常量和配置
+│   └── constants.py   # Agent 配置、显示名称等
+├── core/             # Pipeline 执行逻辑
+├── data/             # 市场数据处理
+│   ├── provider_router.py  # 数据源路由
+│   └── schema.py          # 数据 schema
+├── enrich/           # LLM 响应富化
+├── explain/          # 交易决策解释
+├── llm/              # LLM 集成
+│   └── models.py     # RetryChatModel、TokenRecordingModelWrapper
+├── skills/           # 技能定义（内置 + 自定义）
+├── tools/            # 交易和分析工具
+└── utils/            # 工具函数
+```
+
+## 前端结构
+
+```
+frontend/src/
+├── App.jsx              # React 主应用
+├── components/          # React 组件
+│   ├── RuntimeView.jsx           # 交易运行时 UI
+│   ├── TraderView.jsx            # 交易员界面
+│   ├── RoomView.jsx              # 聊天室视图
+│   ├── StockExplainView.jsx      # 股票解释视图
+│   ├── RuntimeSettingsPanel.jsx  # 运行时设置面板
+│   ├── WatchlistPanel.jsx        # 关注列表
+│   ├── PerformanceView.jsx       # 绩效视图
+│   ├── StatisticsView.jsx        # 统计视图
+│   ├── NetValueChart.jsx         # 净值曲线图
+│   ├── AgentCard.jsx             # Agent 卡片
+│   ├── AgentFeed.jsx             # Agent 动态
+│   └── explain/                  # 解释相关组件
+│       ├── ExplainNewsSection.jsx
+│       ├── ExplainRangeSection.jsx
+│       ├── ExplainSimilarDaysSection.jsx
+│       ├── ExplainStorySection.jsx
+│       └── useExplainModel.js
+├── services/            # API 服务
+│   ├── runtimeApi.js     # 运行时 API 调用
+│   ├── websocket.js      # WebSocket 实时通信
+│   ├── newsApi.js        # 新闻服务客户端
+│   └── tradingApi.js     # 交易服务客户端
+├── config/
+│   └── constants.js      # Agent 定义、配置
+└── hooks/               # React Hooks
+```
+
+## Agent 系统
+
+### 6 种 Agent 角色
+
+| 角色 ID | 名称 | 职责 |
+|---------|------|------|
+| `fundamentals_analyst` | 基本面分析师 | 财务健康、盈利能力、成长质量 |
+| `technical_analyst` | 技术分析师 | 价格趋势、技术指标、动量分析 |
+| `sentiment_analyst` | 情绪分析师 | 市场情绪、新闻情绪、内幕交易 |
+| `valuation_analyst` | 估值分析师 | DCF、EV/EBITDA、 intrinsic value |
+| `portfolio_manager` | 投资经理 | 决策执行、交易协调 |
+| `risk_manager` | 风控经理 | 实时价格/波动率监控、仓位限制、多层风险预警 |
+
+### Hook 系统 (`base/hooks.py`)
+
+- **MemoryCompactionHook**: 基于 CoPaw 的内存压缩
+  - `memory_compact_ratio`: 压缩目标比例（默认 0.75）
+  - `memory_reserve_ratio`: 保留比例（默认 0.1）
+  - `enable_tool_result_compact`: 工具结果压缩
+  - `tool_result_compact_keep_n`: 保留最近 N 条工具结果
+
+### 添加自定义分析师
+
+1. 在 `backend/agents/prompts/analyst/personas.yaml` 注册
+2. 在 `backend/config/constants.py` 的 `ANALYST_TYPES` 字典中添加
+3. 可选：在 `frontend/src/config/constants.js` 中更新前端配置
+
+### LLM 模型封装 (`backend/llm/models.py`)
+
+基于 CoPaw 的模型封装设计：
+
+- **RetryChatModel**: 自动重试瞬态 LLM 错误（rate limit、timeout、502/503 等），指数退避
+  - `max_retries`: 最大重试次数（默认 3）
+  - `initial_delay`: 初始延迟秒数（默认 1.0）
+  - `backoff_multiplier`: 退避倍数（默认 2.0）
+
+- **TokenRecordingModelWrapper**: 追踪每个 provider 的 token 消耗和成本
+
+```python
+from backend.llm.models import create_model, RetryChatModel
+
+model = RetryChatModel(create_model("gpt-4o", "OPENAI"), max_retries=3)
+```
+
+## 技能系统 (`backend/skills/`)
+
+技能定义在 `SKILL.md` 文件中，包含：
+- `instructions` - 技能说明
+- `triggers` - 触发条件
+- `parameters` - 输入/输出 schema
+- `available_tools` - 技能可使用的工具
+
+技能由 `skills_manager.py` 加载，通过 `skill_adaptation_hook.py` 绑定到 Agent。
+
+技能管理器支持 6 种作用域：builtin、customized、installed、active、disabled、local。
+
+## Pipeline 执行 (`backend/core/`)
+
+每日交易流程：
+
+1. **分析阶段** - 各 Agent 基于工具和历史经验独立分析
+2. **沟通阶段** - 通过私聊、通知、会议等方式交换观点（1v1/1vN/NvN）
+3. **决策阶段** - 投资经理综合判断，给出最终交易
+4. **评估阶段** - 绩效跟踪
+5. **复盘阶段** - Agent 根据当日实际收益反思总结，通过 ReMe 记忆框架更新经验
+
+## 前端状态管理
+
+项目正在向 Zustand 状态管理过渡，已创建的 store：
+
+```bash
+frontend/src/store/
+├── index.js           # 导出所有 store
+├── runtimeStore.js    # 连接状态、运行时配置
+├── marketStore.js     # 市场数据、股票价格
+├── portfolioStore.js  # 组合、持仓、交易
+├── agentStore.js      # Agent 技能、工作区
+└── uiStore.js         # UI 状态、视图切换
+```
+
+**迁移状态**：
+- Stores 已创建但尚未在 App.jsx 中使用
+- 计划：逐步迁移 60+ 个 useState 到对应 store
+
+## 环境配置
+
+`.env` 必需配置：
+
+```bash
+# 金融数据源
+FIN_DATA_SOURCE=finnhub|financial_datasets
+FINANCIAL_DATASETS_API_KEY=   # 回测必需
+FINNHUB_API_KEY=              # 实盘必需
+
+# Agent LLM
+OPENAI_API_KEY=
+OPENAI_BASE_URL=
+MODEL_NAME=qwen3-max-preview
+
+# 可为不同 Agent 指定不同模型
+AGENT_SENTIMENT_ANALYST_MODEL_NAME=qwen3-max-preview
+AGENT_FUNDAMENTALS_ANALYST_MODEL_NAME=deepseek-chat
+
+# ReMe 记忆系统
+MEMORY_API_KEY=
+```
+
+## 关键依赖
+
+- **AgentScope** - 多智能体框架
+- **ReMe** - 持续学习记忆系统
+- **FastAPI** + **uvicorn** - 后端 API 服务器
+- **websockets** - 实时通信
+- **React 19** + **Vite** + **TailwindCSS** - 前端
+- **React Context** - 前端状态管理（App.jsx 中使用 useState + useCallback）
+- **Three.js** / **React-Three-Fiber** - 3D 可视化
--- a/README_zh.md
+++ b/README_zh.md
@@ -96,8 +96,11 @@ evotraders live                    # 立即运行（默认）
 evotraders live --enable-memory    # 使用记忆
 evotraders live --mock             # Mock 模式（测试）
 evotraders live -t 22:30           # 每天本地时间 22:30 运行（自动转换为 NYSE 时区）
+evotraders live --schedule-mode intraday --interval-minutes 60  # 每隔 1 小时触发一次；仅交易时段执行交易，其他时段只分析
 ```

+前端的“运行设置”面板也支持热更新 `schedule_mode`、`interval_minutes`、`max_comm_cycles`；其中 daily 模式时间当前按 NYSE/ET 配置。
+
 **获取帮助：**
 ```bash
 evotraders --help           # 查看整体命令行帮助
@@ -114,6 +117,54 @@ evotraders frontend                # 默认连接 8765 端口, 你可以修改 .

 访问 `http://localhost:5173/` 查看交易大厅，选择日期并点击 Run/Replay 观察决策过程。

+### 迁移期服务边界说明
+
+当前仓库正处于从模块化单体向独立服务迁移的阶段，当前默认开发路径已经切到独立 app surface：
+
+- `backend.apps.agent_service`
+- `backend.apps.runtime_service`
+- `backend.apps.trading_service`
+- `backend.apps.news_service`
+
+当前本地开发默认推荐直接运行拆分后的服务：
+
+```bash
+./start-dev.sh split
+
+# 或分别手动启动
+python -m uvicorn backend.apps.agent_service:app --port 8000 --reload
+python -m uvicorn backend.apps.runtime_service:app --port 8003 --reload
+python -m uvicorn backend.apps.trading_service:app --port 8001 --reload
+python -m uvicorn backend.apps.news_service:app --port 8002 --reload
+```
+
+迁移期关键环境变量：
+
+```bash
+# 后端 Gateway 优先走独立服务读取
+NEWS_SERVICE_URL=http://localhost:8002
+TRADING_SERVICE_URL=http://localhost:8001
+
+# 前端浏览器直连控制面 / 运行时面
+VITE_CONTROL_API_BASE_URL=http://localhost:8000/api
+VITE_RUNTIME_API_BASE_URL=http://localhost:8003/api/runtime
+
+# 前端浏览器优先直连独立服务
+VITE_NEWS_SERVICE_URL=http://localhost:8002
+VITE_TRADING_SERVICE_URL=http://localhost:8001
+```
+
+目前前端已支持直连 `news-service` 的 explain 只读路径包括：
+
+- runtime panel / gateway port 查询已可独立指向 `runtime-service`
+- story
+- similar days
+- range explain
+- news for date
+- news categories
+
+如果没有配置这些变量，系统会继续走当前保留的本地回退逻辑。
+
 ---

 ## 系统架构
--- a/backend/agents/init.py
+++ b/backend/agents/init.py
@@ -1,6 +1,57 @@
 # -*- coding: utf-8 -*-
+"""
+Agents package - EvoAgent architecture for trading system.
+
+Exports:
+- EvoAgent: Next-generation agent with workspace support
+- ToolGuardMixin: Tool call approval/denial flow
+- CommandHandler: System command handling
+- AgentFactory: Dynamic agent creation and management
+- WorkspaceManager: Legacy name for the persistent workspace registry
+- WorkspaceRegistry: Explicit run-time-agnostic workspace registry
+- RunWorkspaceManager: Run-scoped workspace asset manager
+- AgentRegistry: Central agent registry
+- Legacy compatibility: AnalystAgent, PMAgent, RiskAgent
+"""
+
+# New EvoAgent architecture (from agent_core.py)
+from .agent_core import EvoAgent, ToolGuardMixin, CommandHandler
+from .factory import AgentFactory, ModelConfig, RoleConfig
+from .workspace import WorkspaceManager, WorkspaceRegistry, WorkspaceConfig
+from .workspace_manager import RunWorkspaceManager
+from .registry import AgentRegistry, AgentInfo, get_registry, reset_registry
+
+# Legacy agents (backward compatibility)
 from .analyst import AnalystAgent
 from .portfolio_manager import PMAgent
 from .risk_manager import RiskAgent

-__all__ = ["AnalystAgent", "PMAgent", "RiskAgent"]
+# Compatibility layer
+from .compat import LegacyAgentAdapter, adapt_agent, adapt_agents, is_legacy_agent
+
+__all__ = [
+    # New architecture
+    "EvoAgent",
+    "ToolGuardMixin",
+    "CommandHandler",
+    "AgentFactory",
+    "ModelConfig",
+    "RoleConfig",
+    "WorkspaceManager",
+    "WorkspaceRegistry",
+    "WorkspaceConfig",
+    "RunWorkspaceManager",
+    "AgentRegistry",
+    "AgentInfo",
+    "get_registry",
+    "reset_registry",
+    # Legacy compatibility
+    "AnalystAgent",
+    "PMAgent",
+    "RiskAgent",
+    # Compatibility layer
+    "LegacyAgentAdapter",
+    "adapt_agent",
+    "adapt_agents",
+    "is_legacy_agent",
+]
--- a/backend/agents/agent_core.py
+++ b/backend/agents/agent_core.py
@@ -0,0 +1,18 @@
+# -*- coding: utf-8 -*-
+"""
+Compatibility layer for legacy imports.
+
+This module re-exports the newer base implementations so existing import
+paths (`from backend.agents.agent_core import EvoAgent`) continue to work while
+centralizing the actual logic in `backend.agents.base.evo_agent`.
+"""
+
+from .base.command_handler import CommandHandler
+from .base.evo_agent import EvoAgent
+from .base.tool_guard import ToolGuardMixin
+
+__all__ = [
+    "EvoAgent",
+    "ToolGuardMixin",
+    "CommandHandler",
+]
--- a/backend/agents/agent_workspace.py
+++ b/backend/agents/agent_workspace.py
@@ -0,0 +1,75 @@
+# -*- coding: utf-8 -*-
+"""Per-agent run-scoped workspace configuration helpers."""
+
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+import yaml
+
+
+@dataclass(frozen=True)
+class AgentWorkspaceConfig:
+    """Structured agent config loaded from runs/<config>/agents/<agent>/agent.yaml."""
+
+    values: Dict[str, Any] = field(default_factory=dict)
+
+    @property
+    def prompt_files(self) -> Optional[List[str]]:
+        raw = self.values.get("prompt_files")
+        if not isinstance(raw, list):
+            return None
+        files = [
+            str(item).strip()
+            for item in raw
+            if isinstance(item, str) and str(item).strip()
+        ]
+        return files or None
+
+    @property
+    def enabled_skills(self) -> List[str]:
+        return _normalized_string_list(self.values.get("enabled_skills"))
+
+    @property
+    def disabled_skills(self) -> List[str]:
+        return _normalized_string_list(self.values.get("disabled_skills"))
+
+    @property
+    def active_tool_groups(self) -> Optional[List[str]]:
+        groups = _normalized_string_list(self.values.get("active_tool_groups"))
+        return groups or None
+
+    @property
+    def disabled_tool_groups(self) -> List[str]:
+        return _normalized_string_list(self.values.get("disabled_tool_groups"))
+
+    def get(self, key: str, default: Any = None) -> Any:
+        return self.values.get(key, default)
+
+
+def _normalized_string_list(raw: Any) -> List[str]:
+    if not isinstance(raw, list):
+        return []
+    seen: List[str] = []
+    for item in raw:
+        if not isinstance(item, str):
+            continue
+        value = item.strip()
+        if value and value not in seen:
+            seen.append(value)
+    return seen
+
+
+def load_agent_workspace_config(path: Path) -> AgentWorkspaceConfig:
+    """Load agent.yaml if present."""
+    if not path.exists() or not path.is_file():
+        return AgentWorkspaceConfig()
+
+    raw = path.read_text(encoding="utf-8").strip()
+    if not raw:
+        return AgentWorkspaceConfig()
+
+    parsed = yaml.safe_load(raw) or {}
+    if not isinstance(parsed, dict):
+        parsed = {}
+    return AgentWorkspaceConfig(values=parsed)
--- a/backend/agents/analyst.py
+++ b/backend/agents/analyst.py
@@ -11,9 +11,7 @@ from agentscope.message import Msg

 from ..config.constants import ANALYST_TYPES
 from ..utils.progress import progress
-from .prompt_loader import PromptLoader
-
-_prompt_loader = PromptLoader()
+from .prompt_factory import build_agent_system_prompt, clear_prompt_factory_cache


 class AnalystAgent(ReActAgent):
@@ -50,14 +48,19 @@ class AnalystAgent(ReActAgent):
                f"Must be one of: {list(ANALYST_TYPES.keys())}",
            )

-        self.analyst_type_key = analyst_type
-        self.analyst_persona = ANALYST_TYPES[analyst_type]["display_name"]
+        object.__setattr__(self, "analyst_type_key", analyst_type)
+        object.__setattr__(
+            self,
+            "analyst_persona",
+            ANALYST_TYPES[analyst_type]["display_name"],
+        )

        if agent_id is None:
            agent_id = analyst_type
+        object.__setattr__(self, "agent_id", agent_id)

-        self.config = config or {}
-
+        object.__setattr__(self, "config", config or {})
+        object.__setattr__(self, "toolkit", toolkit)
        sys_prompt = self._load_system_prompt()

        kwargs = {
@@ -77,27 +80,11 @@ class AnalystAgent(ReActAgent):

    def _load_system_prompt(self) -> str:
        """Load system prompt for analyst"""
-        personas_config = _prompt_loader.load_yaml_config(
-            "analyst",
-            "personas",
-        )
-        persona = personas_config.get(self.analyst_type_key, {})
-
-        # Get focus items and format as bullet points
-        focus_items = persona.get("focus", [])
-        focus_text = "\n".join(f"- {item}" for item in focus_items)
-
-        # Get description
-        description = persona.get("description", "").strip()
-
-        return _prompt_loader.load_prompt(
-            "analyst",
-            "system",
-            variables={
-                "analyst_type": self.analyst_persona,
-                "focus": focus_text,
-                "description": description,
-            },
+        return build_agent_system_prompt(
+            agent_id=self.agent_id,
+            config_name=self.config.get("config_name", "default"),
+            toolkit=self.toolkit,
+            analyst_type=self.analyst_type_key,
        )

    async def reply(self, x: Msg = None) -> Msg:
@@ -131,3 +118,23 @@ class AnalystAgent(ReActAgent):
            )

        return result
+
+    def reload_runtime_assets(self, active_skill_dirs: Optional[list] = None) -> None:
+        """Reload toolkit and system prompt from current run assets."""
+        from .toolkit_factory import create_agent_toolkit
+
+        clear_prompt_factory_cache()
+        self.toolkit = create_agent_toolkit(
+            self.agent_id,
+            self.config.get("config_name", "default"),
+            active_skill_dirs=active_skill_dirs,
+        )
+        self._apply_runtime_sys_prompt(self._load_system_prompt())
+
+    def _apply_runtime_sys_prompt(self, sys_prompt: str) -> None:
+        """Update the prompt used by future turns and the cached system msg."""
+        self._sys_prompt = sys_prompt
+        for msg, _marks in self.memory.content:
+            if getattr(msg, "role", None) == "system":
+                msg.content = sys_prompt
+                break
--- a/backend/agents/base/init.py
+++ b/backend/agents/base/init.py
@@ -0,0 +1,57 @@
+# -*- coding: utf-8 -*-
+"""Base agent module for EvoTraders.
+
+提供Agent基础类、命令处理、工具守卫和钩子管理等功能。
+"""
+
+# 命令处理器 (从command_handler.py导入)
+from .command_handler import (
+    AgentCommandDispatcher,
+    CommandContext,
+    CommandHandler,
+    CommandResult,
+    create_command_dispatcher,
+)
+
+# 评估钩子 (从evaluation_hook.py导入)
+from .evaluation_hook import (
+    EvaluationHook,
+    EvaluationCollector,
+    MetricType,
+    EvaluationMetric,
+    EvaluationResult,
+    parse_evaluation_hooks,
+)
+
+# 技能适配钩子 (从skill_adaptation_hook.py导入)
+from .skill_adaptation_hook import (
+    AdaptationAction,
+    AdaptationThreshold,
+    AdaptationEvent,
+    SkillAdaptationHook,
+    AdaptationManager,
+    get_adaptation_manager,
+)
+
+__all__ = [
+    # 命令处理
+    "AgentCommandDispatcher",
+    "CommandContext",
+    "CommandHandler",
+    "CommandResult",
+    "create_command_dispatcher",
+    # 评估钩子
+    "EvaluationHook",
+    "EvaluationCollector",
+    "MetricType",
+    "EvaluationMetric",
+    "EvaluationResult",
+    "parse_evaluation_hooks",
+    # 技能适配钩子
+    "AdaptationAction",
+    "AdaptationThreshold",
+    "AdaptationEvent",
+    "SkillAdaptationHook",
+    "AdaptationManager",
+    "get_adaptation_manager",
+]
--- a/backend/agents/base/command_handler.py
+++ b/backend/agents/base/command_handler.py
@@ -0,0 +1,543 @@
+# -*- coding: utf-8 -*-
+"""Agent command handler for system commands.
+
+This module handles system commands like /save, /compact, /skills, /reload, etc.
+参考CoPaw设计，为EvoAgent提供命令处理能力。
+"""
+import logging
+from abc import ABC, abstractmethod
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import TYPE_CHECKING, Any, Callable, Dict, List, Optional, Protocol
+
+if TYPE_CHECKING:
+    from .agent import EvoAgent
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class CommandResult:
+    """命令执行结果"""
+    success: bool
+    message: str
+    data: Dict[str, Any] = field(default_factory=dict)
+
+
+class CommandContext:
+    """命令执行上下文"""
+
+    def __init__(self, agent: "EvoAgent", raw_query: str, args: str = ""):
+        self.agent = agent
+        self.raw_query = raw_query
+        self.args = args
+        self.config_name = getattr(agent, "config_name", "default")
+        self.agent_id = getattr(agent, "agent_id", "unknown")
+
+
+class CommandHandler(ABC):
+    """命令处理器抽象基类"""
+
+    @abstractmethod
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        """处理命令"""
+        pass
+
+
+class SaveCommandHandler(CommandHandler):
+    """处理 /save <message> 命令 - 保存内容到MEMORY.md"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        message = ctx.args.strip()
+        if not message:
+            return CommandResult(
+                success=False,
+                message="Usage: /save <message>\n请提供要保存的内容。"
+            )
+
+        try:
+            memory_path = self._get_memory_path(ctx)
+            memory_path.parent.mkdir(parents=True, exist_ok=True)
+
+            timestamp = self._get_timestamp()
+            entry = f"\n## {timestamp}\n\n{message}\n"
+
+            with open(memory_path, "a", encoding="utf-8") as f:
+                f.write(entry)
+
+            return CommandResult(
+                success=True,
+                message=f"✅ 内容已保存到 MEMORY.md\n- 路径: {memory_path}\n- 长度: {len(message)} 字符",
+                data={"path": str(memory_path), "length": len(message)}
+            )
+        except Exception as e:
+            logger.error(f"Failed to save to MEMORY.md: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 保存失败: {str(e)}"
+            )
+
+    def _get_memory_path(self, ctx: CommandContext) -> Path:
+        """获取MEMORY.md路径"""
+        from backend.agents.skills_manager import SkillsManager
+        sm = SkillsManager()
+        asset_dir = sm.get_agent_asset_dir(ctx.config_name, ctx.agent_id)
+        return asset_dir / "MEMORY.md"
+
+    def _get_timestamp(self) -> str:
+        """获取当前时间戳"""
+        from datetime import datetime
+        return datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+
+
+class CompactCommandHandler(CommandHandler):
+    """处理 /compact 命令 - 压缩记忆"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        try:
+            agent = ctx.agent
+            memory_manager = getattr(agent, "memory_manager", None)
+
+            if memory_manager is None:
+                return CommandResult(
+                    success=False,
+                    message="❌ Memory Manager 未启用\n\n- 记忆压缩功能不可用\n- 请在配置中启用 memory_manager"
+                )
+
+            messages = await self._get_messages(agent)
+            if not messages:
+                return CommandResult(
+                    success=False,
+                    message="⚠️ 没有可压缩的消息\n\n- 当前记忆为空\n- 无需执行压缩"
+                )
+
+            compact_content = await memory_manager.compact_memory(messages)
+            await self._update_compressed_summary(agent, compact_content)
+
+            return CommandResult(
+                success=True,
+                message=f"✅ 记忆压缩完成\n\n- 压缩了 {len(messages)} 条消息\n- 摘要长度: {len(compact_content)} 字符",
+                data={"message_count": len(messages), "summary_length": len(compact_content)}
+            )
+        except Exception as e:
+            logger.error(f"Failed to compact memory: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 压缩失败: {str(e)}"
+            )
+
+    async def _get_messages(self, agent: "EvoAgent") -> List[Any]:
+        """获取Agent的记忆消息"""
+        memory = getattr(agent, "memory", None)
+        if memory is None:
+            return []
+        return await memory.get_memory() if hasattr(memory, "get_memory") else []
+
+    async def _update_compressed_summary(self, agent: "EvoAgent", content: str) -> None:
+        """更新压缩摘要"""
+        memory = getattr(agent, "memory", None)
+        if memory and hasattr(memory, "update_compressed_summary"):
+            await memory.update_compressed_summary(content)
+
+
+class SkillsListCommandHandler(CommandHandler):
+    """处理 /skills list 命令 - 列出已激活技能"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        try:
+            from backend.agents.skills_manager import SkillsManager
+            sm = SkillsManager()
+
+            active_skills = sm.list_active_skill_metadata(ctx.config_name, ctx.agent_id)
+            catalog = sm.list_agent_skill_catalog(ctx.config_name, ctx.agent_id)
+
+            lines = ["📋 技能列表", ""]
+
+            if active_skills:
+                lines.append("✅ 已激活技能:")
+                for skill in active_skills:
+                    lines.append(f"  • {skill.name} - {skill.description[:50]}...")
+            else:
+                lines.append("⚠️ 当前没有激活的技能")
+
+            lines.append("")
+            lines.append(f"📚 可用技能总数: {len(catalog)}")
+            lines.append("💡 使用 /skills enable <name> 启用技能")
+
+            return CommandResult(
+                success=True,
+                message="\n".join(lines),
+                data={
+                    "active_count": len(active_skills),
+                    "catalog_count": len(catalog),
+                    "active": [s.skill_name for s in active_skills]
+                }
+            )
+        except Exception as e:
+            logger.error(f"Failed to list skills: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 获取技能列表失败: {str(e)}"
+            )
+
+
+class SkillsEnableCommandHandler(CommandHandler):
+    """处理 /skills enable <name> 命令 - 启用技能"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        skill_name = ctx.args.strip()
+        if not skill_name:
+            return CommandResult(
+                success=False,
+                message="Usage: /skills enable <skill_name>\n请提供技能名称。"
+            )
+
+        try:
+            from backend.agents.skills_manager import SkillsManager
+            sm = SkillsManager()
+
+            result = sm.update_agent_skill_overrides(
+                ctx.config_name,
+                ctx.agent_id,
+                enable=[skill_name]
+            )
+
+            return CommandResult(
+                success=True,
+                message=f"✅ 技能已启用: {skill_name}\n\n已启用技能: {', '.join(result['enabled_skills'])}",
+                data=result
+            )
+        except Exception as e:
+            logger.error(f"Failed to enable skill: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 启用技能失败: {str(e)}"
+            )
+
+
+class SkillsDisableCommandHandler(CommandHandler):
+    """处理 /skills disable <name> 命令 - 禁用技能"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        skill_name = ctx.args.strip()
+        if not skill_name:
+            return CommandResult(
+                success=False,
+                message="Usage: /skills disable <skill_name>\n请提供技能名称。"
+            )
+
+        try:
+            from backend.agents.skills_manager import SkillsManager
+            sm = SkillsManager()
+
+            result = sm.update_agent_skill_overrides(
+                ctx.config_name,
+                ctx.agent_id,
+                disable=[skill_name]
+            )
+
+            return CommandResult(
+                success=True,
+                message=f"✅ 技能已禁用: {skill_name}\n\n已禁用技能: {', '.join(result['disabled_skills'])}",
+                data=result
+            )
+        except Exception as e:
+            logger.error(f"Failed to disable skill: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 禁用技能失败: {str(e)}"
+            )
+
+
+class SkillsInstallCommandHandler(CommandHandler):
+    """处理 /skills install <name> 命令 - 安装技能"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        skill_name = ctx.args.strip()
+        if not skill_name:
+            return CommandResult(
+                success=False,
+                message="Usage: /skills install <skill_name>\n请提供技能名称。"
+            )
+
+        try:
+            from backend.agents.skills_manager import SkillsManager
+            from backend.agents.skill_loader import load_skill_from_dir
+            sm = SkillsManager()
+
+            # 查找技能源目录
+            source_dir = self._resolve_skill_source(sm, skill_name)
+            if not source_dir:
+                return CommandResult(
+                    success=False,
+                    message=f"❌ 技能未找到: {skill_name}\n\n请检查技能名称是否正确，或技能是否存在于 builtin/customized 目录。"
+                )
+
+            # 加载并验证技能
+            skill_info = load_skill_from_dir(source_dir)
+            if not skill_info:
+                return CommandResult(
+                    success=False,
+                    message=f"❌ 技能加载失败: {skill_name}\n\n技能格式可能不正确。"
+                )
+
+            # 安装到agent的installed目录
+            installed_root = sm.get_agent_installed_root(ctx.config_name, ctx.agent_id)
+            target_dir = installed_root / skill_name
+
+            import shutil
+            if target_dir.exists():
+                shutil.rmtree(target_dir)
+            shutil.copytree(source_dir, target_dir)
+
+            return CommandResult(
+                success=True,
+                message=f"✅ 技能已安装: {skill_name}\n\n- 名称: {skill_info.get('name', skill_name)}\n- 版本: {skill_info.get('version', 'unknown')}\n- 路径: {target_dir}",
+                data={"skill_name": skill_name, "target_dir": str(target_dir)}
+            )
+        except Exception as e:
+            logger.error(f"Failed to install skill: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 安装技能失败: {str(e)}"
+            )
+
+    def _resolve_skill_source(self, sm: "SkillsManager", skill_name: str) -> Optional[Path]:
+        """解析技能源目录"""
+        for root in [sm.customized_root, sm.builtin_root]:
+            candidate = root / skill_name
+            if candidate.exists() and (candidate / "SKILL.md").exists():
+                return candidate
+        return None
+
+
+class ReloadCommandHandler(CommandHandler):
+    """处理 /reload 命令 - 重新加载配置"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        try:
+            agent = ctx.agent
+
+            # 重新加载配置
+            if hasattr(agent, "reload_config"):
+                await agent.reload_config()
+
+            # 重新加载技能
+            from backend.agents.skills_manager import SkillsManager
+            sm = SkillsManager()
+
+            # 刷新技能同步
+            active_root = sm.get_agent_active_root(ctx.config_name, ctx.agent_id)
+            if active_root.exists():
+                # 清除缓存，强制重新加载
+                import shutil
+                for item in active_root.iterdir():
+                    if item.is_dir():
+                        shutil.rmtree(item)
+
+            return CommandResult(
+                success=True,
+                message="✅ 配置已重新加载\n\n- Agent配置已刷新\n- 技能缓存已清除\n- 请重启对话以应用所有更改",
+                data={"config_name": ctx.config_name, "agent_id": ctx.agent_id}
+            )
+        except Exception as e:
+            logger.error(f"Failed to reload config: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 重新加载失败: {str(e)}"
+            )
+
+
+class StatusCommandHandler(CommandHandler):
+    """处理 /status 命令 - 显示Agent状态"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        try:
+            agent = ctx.agent
+
+            lines = ["📊 Agent 状态", ""]
+            lines.append(f"🆔 Agent ID: {ctx.agent_id}")
+            lines.append(f"⚙️  Config: {ctx.config_name}")
+
+            # 模型信息
+            model = getattr(agent, "model", None)
+            if model:
+                lines.append(f"🤖 Model: {model}")
+
+            # 记忆状态
+            memory = getattr(agent, "memory", None)
+            if memory:
+                msg_count = len(getattr(memory, "content", []))
+                lines.append(f"💾 Memory: {msg_count} messages")
+
+            # 技能状态
+            from backend.agents.skills_manager import SkillsManager
+            sm = SkillsManager()
+            active_skills = sm.list_active_skill_metadata(ctx.config_name, ctx.agent_id)
+            lines.append(f"🔧 Active Skills: {len(active_skills)}")
+
+            # 工具组状态
+            toolkit = getattr(agent, "toolkit", None)
+            if toolkit:
+                groups = getattr(toolkit, "tool_groups", {})
+                active_groups = [name for name, g in groups.items() if getattr(g, "active", False)]
+                lines.append(f"🛠️  Active Tool Groups: {', '.join(active_groups) if active_groups else 'None'}")
+
+            return CommandResult(
+                success=True,
+                message="\n".join(lines),
+                data={
+                    "agent_id": ctx.agent_id,
+                    "config_name": ctx.config_name,
+                    "active_skills_count": len(active_skills)
+                }
+            )
+        except Exception as e:
+            logger.error(f"Failed to get status: {e}")
+            return CommandResult(
+                success=False,
+                message=f"❌ 获取状态失败: {str(e)}"
+            )
+
+
+class HelpCommandHandler(CommandHandler):
+    """处理 /help 命令 - 显示帮助"""
+
+    async def handle(self, ctx: CommandContext) -> CommandResult:
+        help_text = """📖 EvoAgent 命令帮助
+
+可用命令:
+  /save <message>       - 保存内容到 MEMORY.md
+  /compact              - 压缩记忆
+  /skills list          - 列出已激活技能
+  /skills enable <name> - 启用技能
+  /skills disable <name>- 禁用技能
+  /skills install <name>- 安装技能
+  /reload               - 重新加载配置
+  /status               - 显示Agent状态
+  /help                 - 显示此帮助信息
+
+提示:
+  • 所有命令以 / 开头
+  • 命令不区分大小写
+  • 使用 Tab 键可自动补全命令
+"""
+        return CommandResult(success=True, message=help_text)
+
+
+class AgentCommandDispatcher:
+    """Agent命令分发器
+
+    参考CoPaw的CommandHandler设计，为EvoAgent提供统一的命令处理入口。
+    """
+
+    # 支持的系统命令
+    SYSTEM_COMMANDS = frozenset({
+        "save", "compact",
+        "skills", "reload",
+        "status", "help"
+    })
+
+    def __init__(self):
+        self._handlers: Dict[str, CommandHandler] = {}
+        self._subcommands: Dict[str, Dict[str, CommandHandler]] = {}
+        self._register_default_handlers()
+
+    def _register_default_handlers(self) -> None:
+        """注册默认命令处理器"""
+        self._handlers["save"] = SaveCommandHandler()
+        self._handlers["compact"] = CompactCommandHandler()
+        self._handlers["reload"] = ReloadCommandHandler()
+        self._handlers["status"] = StatusCommandHandler()
+        self._handlers["help"] = HelpCommandHandler()
+
+        # 子命令: /skills list/enable/disable/install
+        self._subcommands["skills"] = {
+            "list": SkillsListCommandHandler(),
+            "enable": SkillsEnableCommandHandler(),
+            "disable": SkillsDisableCommandHandler(),
+            "install": SkillsInstallCommandHandler(),
+        }
+
+    def is_command(self, query: str | None) -> bool:
+        """检查是否为命令
+
+        Args:
+            query: 用户输入字符串
+
+        Returns:
+            True 如果是系统命令
+        """
+        if not isinstance(query, str) or not query.startswith("/"):
+            return False
+
+        parts = query.strip().lstrip("/").split()
+        if not parts:
+            return False
+
+        cmd = parts[0].lower()
+
+        # 检查主命令
+        if cmd in self.SYSTEM_COMMANDS:
+            return True
+
+        return False
+
+    async def handle(self, agent: "EvoAgent", query: str) -> CommandResult:
+        """处理命令
+
+        Args:
+            agent: EvoAgent实例
+            query: 命令字符串
+
+        Returns:
+            命令执行结果
+        """
+        if not self.is_command(query):
+            return CommandResult(
+                success=False,
+                message=f"未知命令: {query}\n使用 /help 查看可用命令。"
+            )
+
+        # 解析命令和参数
+        parts = query.strip().lstrip("/").split(maxsplit=1)
+        cmd = parts[0].lower()
+        args = parts[1] if len(parts) > 1 else ""
+
+        logger.info(f"Processing command: {cmd}, args: {args}")
+
+        # 处理子命令 (e.g., /skills list)
+        if cmd in self._subcommands:
+            sub_parts = args.split(maxsplit=1)
+            sub_cmd = sub_parts[0].lower() if sub_parts else ""
+            sub_args = sub_parts[1] if len(sub_parts) > 1 else ""
+
+            handlers = self._subcommands[cmd]
+            handler = handlers.get(sub_cmd)
+
+            if handler is None:
+                available = ", ".join(handlers.keys())
+                return CommandResult(
+                    success=False,
+                    message=f"未知子命令: {sub_cmd}\n可用子命令: {available}"
+                )
+
+            ctx = CommandContext(agent, query, sub_args)
+            return await handler.handle(ctx)
+
+        # 处理主命令
+        handler = self._handlers.get(cmd)
+        if handler is None:
+            return CommandResult(
+                success=False,
+                message=f"命令未实现: {cmd}"
+            )
+
+        ctx = CommandContext(agent, query, args)
+        return await handler.handle(ctx)
+
+
+# 便捷函数
+def create_command_dispatcher() -> AgentCommandDispatcher:
+    """创建命令分发器实例"""
+    return AgentCommandDispatcher()
--- a/backend/agents/base/evaluation_hook.py
+++ b/backend/agents/base/evaluation_hook.py
@@ -0,0 +1,452 @@
+# -*- coding: utf-8 -*-
+"""Evaluation hooks system for skills.
+
+Provides evaluation metric collection and storage for skill performance tracking.
+Based on the evaluation hooks design in SKILL_TEMPLATE.md.
+"""
+from __future__ import annotations
+
+import json
+import logging
+from dataclasses import dataclass, field, asdict
+from datetime import datetime
+from enum import Enum
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Set
+
+logger = logging.getLogger(__name__)
+
+
+class MetricType(Enum):
+    """Types of evaluation metrics."""
+    HIT_RATE = "hit_rate"           # 信号命中率
+    RISK_VIOLATION = "risk_violation"  # 风控违例率
+    POSITION_DEVIATION = "position_deviation"  # 仓位偏离率
+    PnL_ATTRIBUTION = "pnl_attribution"  # P&L 归因一致性
+    SIGNAL_CONSISTENCY = "signal_consistency"  # 信号一致性
+    DECISION_LATENCY = "decision_latency"  # 决策延迟
+    TOOL_USAGE = "tool_usage"  # 工具使用率
+    CUSTOM = "custom"  # 自定义指标
+
+
+@dataclass
+class EvaluationMetric:
+    """A single evaluation metric."""
+    name: str
+    metric_type: MetricType
+    value: float
+    timestamp: str = field(default_factory=lambda: datetime.now().isoformat())
+    metadata: Dict[str, Any] = field(default_factory=dict)
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "name": self.name,
+            "metric_type": self.metric_type.value,
+            "value": self.value,
+            "timestamp": self.timestamp,
+            "metadata": self.metadata,
+        }
+
+
+@dataclass
+class EvaluationResult:
+    """Evaluation result for a skill execution."""
+    skill_name: str
+    run_id: str
+    agent_id: str
+    metrics: List[EvaluationMetric] = field(default_factory=list)
+    inputs: Dict[str, Any] = field(default_factory=dict)
+    outputs: Dict[str, Any] = field(default_factory=dict)
+    decision: Optional[str] = None
+    success: bool = True
+    error_message: Optional[str] = None
+    started_at: Optional[str] = None
+    completed_at: Optional[str] = field(default_factory=lambda: datetime.now().isoformat())
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "skill_name": self.skill_name,
+            "run_id": self.run_id,
+            "agent_id": self.agent_id,
+            "metrics": [m.to_dict() for m in self.metrics],
+            "inputs": self.inputs,
+            "outputs": self.outputs,
+            "decision": self.decision,
+            "success": self.success,
+            "error_message": self.error_message,
+            "started_at": self.started_at,
+            "completed_at": self.completed_at,
+        }
+
+
+class EvaluationHook:
+    """Hook for collecting skill evaluation metrics.
+
+    This hook collects and stores evaluation metrics after skill execution
+    for later analysis and memory/reflection stages.
+    """
+
+    def __init__(
+        self,
+        storage_dir: Path,
+        run_id: str,
+        agent_id: str,
+    ):
+        """Initialize evaluation hook.
+
+        Args:
+            storage_dir: Directory to store evaluation results
+            run_id: Current run identifier
+            agent_id: Current agent identifier
+        """
+        self.storage_dir = Path(storage_dir)
+        self.run_id = run_id
+        self.agent_id = agent_id
+        self._current_evaluation: Optional[EvaluationResult] = None
+
+    def start_evaluation(
+        self,
+        skill_name: str,
+        inputs: Dict[str, Any],
+    ) -> None:
+        """Start a new evaluation session.
+
+        Args:
+            skill_name: Name of the skill being evaluated
+            inputs: Input parameters for the skill
+        """
+        self._current_evaluation = EvaluationResult(
+            skill_name=skill_name,
+            run_id=self.run_id,
+            agent_id=self.agent_id,
+            inputs=inputs,
+            started_at=datetime.now().isoformat(),
+        )
+        logger.debug(f"Started evaluation for skill: {skill_name}")
+
+    def add_metric(
+        self,
+        name: str,
+        metric_type: MetricType,
+        value: float,
+        metadata: Optional[Dict[str, Any]] = None,
+    ) -> None:
+        """Add an evaluation metric.
+
+        Args:
+            name: Metric name
+            metric_type: Type of metric
+            value: Metric value
+            metadata: Additional metadata
+        """
+        if self._current_evaluation is None:
+            logger.warning("No active evaluation session, ignoring metric")
+            return
+
+        metric = EvaluationMetric(
+            name=name,
+            metric_type=metric_type,
+            value=value,
+            metadata=metadata or {},
+        )
+        self._current_evaluation.metrics.append(metric)
+        logger.debug(f"Added metric: {name} = {value}")
+
+    def add_metrics(self, metrics: List[EvaluationMetric]) -> None:
+        """Add multiple evaluation metrics at once.
+
+        Args:
+            metrics: List of metrics to add
+        """
+        if self._current_evaluation is None:
+            logger.warning("No active evaluation session, ignoring metrics")
+            return
+
+        self._current_evaluation.metrics.extend(metrics)
+
+    def record_outputs(self, outputs: Dict[str, Any]) -> None:
+        """Record skill outputs.
+
+        Args:
+            outputs: Output from skill execution
+        """
+        if self._current_evaluation is None:
+            logger.warning("No active evaluation session, ignoring outputs")
+            return
+
+        self._current_evaluation.outputs = outputs
+
+    def record_decision(self, decision: str) -> None:
+        """Record the final decision.
+
+        Args:
+            decision: Final decision made by the skill
+        """
+        if self._current_evaluation is None:
+            logger.warning("No active evaluation session, ignoring decision")
+            return
+
+        self._current_evaluation.decision = decision
+
+    def complete_evaluation(
+        self,
+        success: bool = True,
+        error_message: Optional[str] = None,
+    ) -> Optional[EvaluationResult]:
+        """Complete the evaluation session and persist results.
+
+        Args:
+            success: Whether the skill execution was successful
+            error_message: Error message if failed
+
+        Returns:
+            The completed evaluation result, or None if no active evaluation
+        """
+        if self._current_evaluation is None:
+            logger.warning("No active evaluation to complete")
+            return None
+
+        self._current_evaluation.success = success
+        self._current_evaluation.error_message = error_message
+        self._current_evaluation.completed_at = datetime.now().isoformat()
+
+        # Persist to storage
+        result = self._persist_evaluation(self._current_evaluation)
+
+        self._current_evaluation = None
+        logger.debug(f"Completed evaluation for skill: {result.skill_name}")
+
+        return result
+
+    def _persist_evaluation(self, evaluation: EvaluationResult) -> EvaluationResult:
+        """Persist evaluation result to storage.
+
+        Args:
+            evaluation: Evaluation result to persist
+
+        Returns:
+            The persisted evaluation
+        """
+        # Create run-specific directory
+        run_dir = self.storage_dir / self.run_id
+        run_dir.mkdir(parents=True, exist_ok=True)
+
+        # Create agent-specific subdirectory
+        agent_dir = run_dir / self.agent_id
+        agent_dir.mkdir(parents=True, exist_ok=True)
+
+        # Generate filename with timestamp
+        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
+        filename = f"{evaluation.skill_name}_{timestamp}.json"
+        filepath = agent_dir / filename
+
+        # Write evaluation result
+        try:
+            with open(filepath, "w", encoding="utf-8") as f:
+                json.dump(evaluation.to_dict(), f, ensure_ascii=False, indent=2)
+            logger.info(f"Persisted evaluation to: {filepath}")
+        except Exception as e:
+            logger.error(f"Failed to persist evaluation: {e}")
+
+        return evaluation
+
+    def cancel_evaluation(self) -> None:
+        """Cancel the current evaluation session without saving."""
+        if self._current_evaluation is not None:
+            logger.debug(f"Cancelled evaluation for: {self._current_evaluation.skill_name}")
+            self._current_evaluation = None
+
+
+class EvaluationCollector:
+    """Collector for aggregating evaluation metrics across runs.
+
+    Provides methods to query and analyze evaluation results.
+    """
+
+    def __init__(self, storage_dir: Path):
+        """Initialize evaluation collector.
+
+        Args:
+            storage_dir: Root directory containing evaluation results
+        """
+        self.storage_dir = Path(storage_dir)
+
+    def get_run_evaluations(
+        self,
+        run_id: str,
+        agent_id: Optional[str] = None,
+    ) -> List[EvaluationResult]:
+        """Get all evaluations for a run.
+
+        Args:
+            run_id: Run identifier
+            agent_id: Optional agent identifier to filter by
+
+        Returns:
+            List of evaluation results
+        """
+        run_dir = self.storage_dir / run_id
+        if not run_dir.exists():
+            return []
+
+        evaluations = []
+
+        agent_dirs = [run_dir / agent_id] if agent_id else run_dir.iterdir()
+
+        for agent_dir in agent_dirs:
+            if not agent_dir.is_dir():
+                continue
+
+            for eval_file in agent_dir.glob("*.json"):
+                try:
+                    with open(eval_file, "r", encoding="utf-8") as f:
+                        data = json.load(f)
+                        evaluations.append(self._parse_evaluation(data))
+                except Exception as e:
+                    logger.warning(f"Failed to load evaluation {eval_file}: {e}")
+
+        return evaluations
+
+    def get_skill_metrics(
+        self,
+        skill_name: str,
+        run_ids: Optional[List[str]] = None,
+    ) -> List[EvaluationMetric]:
+        """Get all metrics for a specific skill.
+
+        Args:
+            skill_name: Name of the skill
+            run_ids: Optional list of run IDs to filter by
+
+        Returns:
+            List of metrics for the skill
+        """
+        metrics = []
+
+        if run_ids is None:
+            run_ids = [d.name for d in self.storage_dir.iterdir() if d.is_dir()]
+
+        for run_id in run_ids:
+            evaluations = self.get_run_evaluations(run_id)
+            for eval_result in evaluations:
+                if eval_result.skill_name == skill_name:
+                    metrics.extend(eval_result.metrics)
+
+        return metrics
+
+    def calculate_skill_stats(
+        self,
+        skill_name: str,
+        metric_type: MetricType,
+        run_ids: Optional[List[str]] = None,
+    ) -> Dict[str, float]:
+        """Calculate statistics for a specific metric type.
+
+        Args:
+            skill_name: Name of the skill
+            metric_type: Type of metric to calculate
+            run_ids: Optional list of run IDs to filter by
+
+        Returns:
+            Dictionary with min, max, avg, count statistics
+        """
+        metrics = self.get_skill_metrics(skill_name, run_ids)
+        filtered = [m for m in metrics if m.metric_type == metric_type]
+
+        if not filtered:
+            return {"count": 0}
+
+        values = [m.value for m in filtered]
+        return {
+            "count": len(values),
+            "min": min(values),
+            "max": max(values),
+            "avg": sum(values) / len(values),
+        }
+
+    def _parse_evaluation(self, data: Dict[str, Any]) -> EvaluationResult:
+        """Parse evaluation data into EvaluationResult.
+
+        Args:
+            data: Raw evaluation data
+
+        Returns:
+            Parsed EvaluationResult
+        """
+        metrics = []
+        for m in data.get("metrics", []):
+            metrics.append(EvaluationMetric(
+                name=m["name"],
+                metric_type=MetricType(m["metric_type"]),
+                value=m["value"],
+                timestamp=m.get("timestamp", ""),
+                metadata=m.get("metadata", {}),
+            ))
+
+        return EvaluationResult(
+            skill_name=data["skill_name"],
+            run_id=data["run_id"],
+            agent_id=data["agent_id"],
+            metrics=metrics,
+            inputs=data.get("inputs", {}),
+            outputs=data.get("outputs", {}),
+            decision=data.get("decision"),
+            success=data.get("success", True),
+            error_message=data.get("error_message"),
+            started_at=data.get("started_at"),
+            completed_at=data.get("completed_at"),
+        )
+
+
+def parse_evaluation_hooks(skill_dir: Path) -> Dict[str, Any]:
+    """Parse evaluation hooks from SKILL.md.
+
+    Extracts the Optional: Evaluation hooks section from skill documentation.
+
+    Args:
+        skill_dir: Skill directory path
+
+    Returns:
+        Dictionary containing evaluation hook definitions
+    """
+    skill_md = skill_dir / "SKILL.md"
+    if not skill_md.exists():
+        return {}
+
+    try:
+        content = skill_md.read_text(encoding="utf-8")
+
+        # Extract evaluation hooks section
+        if "## Optional: Evaluation hooks" in content:
+            start = content.find("## Optional: Evaluation hooks")
+            # Find the next ## section or end of file
+            next_section = content.find("\n## ", start + 1)
+            if next_section == -1:
+                eval_section = content[start:]
+            else:
+                eval_section = content[start:next_section]
+
+            # Parse metrics from the section
+            metrics = []
+            for metric_type in MetricType:
+                if metric_type.value.replace("_", " ") in eval_section.lower():
+                    metrics.append(metric_type.value)
+
+            return {
+                "supported_metrics": metrics,
+                "section_content": eval_section.strip(),
+            }
+    except Exception as e:
+        logger.warning(f"Failed to parse evaluation hooks: {e}")
+
+    return {}
+
+
+__all__ = [
+    "MetricType",
+    "EvaluationMetric",
+    "EvaluationResult",
+    "EvaluationHook",
+    "EvaluationCollector",
+    "parse_evaluation_hooks",
+]
--- a/backend/agents/base/evo_agent.py
+++ b/backend/agents/base/evo_agent.py
@@ -0,0 +1,510 @@
+# -*- coding: utf-8 -*-
+"""EvoAgent - Core agent implementation for EvoTraders.
+
+This module provides the main EvoAgent class built on AgentScope's ReActAgent,
+with integrated tools, skills, and memory management based on CoPaw design.
+
+Key features:
+- Workspace-driven configuration from Markdown files
+- Dynamic skill loading from skills/active directories
+- Tool-guard security interception
+- Hook system for extensibility
+- Runtime skill and prompt reloading
+"""
+from __future__ import annotations
+
+import logging
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Type, TYPE_CHECKING
+
+from agentscope.agent import ReActAgent
+from agentscope.memory import InMemoryMemory
+from agentscope.message import Msg
+from agentscope.tool import Toolkit
+
+from .tool_guard import ToolGuardMixin
+from .hooks import (
+    HookManager,
+    BootstrapHook,
+    MemoryCompactionHook,
+    WorkspaceWatchHook,
+    HOOK_PRE_REASONING,
+)
+from ..prompts.builder import (
+    PromptBuilder,
+    build_system_prompt_from_workspace,
+)
+from ..agent_workspace import load_agent_workspace_config
+from ..skills_manager import SkillsManager
+
+# Team infrastructure imports (graceful import - may not exist yet)
+try:
+    from backend.agents.team.messenger import AgentMessenger
+    from backend.agents.team.task_delegator import TaskDelegator
+    TEAM_INFRA_AVAILABLE = True
+except ImportError:
+    TEAM_INFRA_AVAILABLE = False
+    AgentMessenger = None
+    TaskDelegator = None
+
+if TYPE_CHECKING:
+    from agentscope.formatter import FormatterBase
+    from agentscope.model import ModelWrapperBase
+
+logger = logging.getLogger(__name__)
+
+
+class EvoAgent(ToolGuardMixin, ReActAgent):
+    """EvoAgent with integrated tools, skills, and memory management.
+
+    This agent extends ReActAgent with:
+    - Workspace-driven configuration from AGENTS.md/SOUL.md/PROFILE.md/etc.
+    - Dynamic skill loading from skills/active directories
+    - Tool-guard security interception (via ToolGuardMixin)
+    - Hook system for extensibility (bootstrap, memory compaction)
+    - Runtime skill and prompt reloading
+
+    MRO note
+    ~~~~~~~~
+    ``ToolGuardMixin`` overrides ``_acting`` and ``_reasoning`` via
+    Python's MRO: EvoAgent → ToolGuardMixin → ReActAgent.
+
+    Example:
+        agent = EvoAgent(
+            agent_id="fundamentals_analyst",
+            config_name="smoke_fullstack",
+            workspace_dir=Path("runs/smoke_fullstack/agents/fundamentals_analyst"),
+            model=model_instance,
+            formatter=formatter_instance,
+        )
+    """
+
+    def __init__(
+        self,
+        agent_id: str,
+        config_name: str,
+        workspace_dir: Path,
+        model: "ModelWrapperBase",
+        formatter: "FormatterBase",
+        skills_manager: Optional[SkillsManager] = None,
+        sys_prompt: Optional[str] = None,
+        max_iters: int = 10,
+        memory: Optional[Any] = None,
+        enable_tool_guard: bool = True,
+        enable_bootstrap_hook: bool = True,
+        enable_memory_compaction: bool = False,
+        memory_manager: Optional[Any] = None,
+        memory_compact_threshold: Optional[int] = None,
+        env_context: Optional[str] = None,
+        prompt_files: Optional[List[str]] = None,
+    ):
+        """Initialize EvoAgent.
+
+        Args:
+            agent_id: Unique identifier for this agent
+            config_name: Run configuration name (e.g., "smoke_fullstack")
+            workspace_dir: Agent workspace directory containing markdown files
+            model: LLM model instance
+            formatter: Message formatter instance
+            skills_manager: Optional SkillsManager instance
+            sys_prompt: Optional override for system prompt
+            max_iters: Maximum reasoning-acting iterations
+            memory: Optional memory instance (defaults to InMemoryMemory)
+            enable_tool_guard: Enable tool-guard security interception
+            enable_bootstrap_hook: Enable bootstrap guidance on first interaction
+            enable_memory_compaction: Enable automatic memory compaction
+            memory_manager: Optional memory manager for compaction
+            memory_compact_threshold: Token threshold for memory compaction
+            env_context: Optional environment context to prepend to system prompt
+            prompt_files: List of markdown files to load (defaults to standard set)
+        """
+        self.agent_id = agent_id
+        self.config_name = config_name
+        self.workspace_dir = Path(workspace_dir)
+        self._skills_manager = skills_manager or SkillsManager()
+        self._env_context = env_context
+        self._prompt_files = prompt_files
+
+        # Initialize tool guard
+        if enable_tool_guard:
+            self._init_tool_guard()
+
+        # Load agent configuration from workspace
+        self._agent_config = self._load_agent_config()
+
+        # Build or use provided system prompt
+        if sys_prompt is not None:
+            self._sys_prompt = sys_prompt
+        else:
+            self._sys_prompt = self._build_system_prompt()
+
+        # Create toolkit with skills
+        toolkit = self._create_toolkit()
+
+        # Initialize hook manager
+        self._hook_manager = HookManager()
+
+        # Initialize parent ReActAgent
+        super().__init__(
+            name=agent_id,
+            model=model,
+            sys_prompt=self._sys_prompt,
+            toolkit=toolkit,
+            memory=memory or InMemoryMemory(),
+            formatter=formatter,
+            max_iters=max_iters,
+        )
+
+        # Register hooks
+        self._register_hooks(
+            enable_bootstrap=enable_bootstrap_hook,
+            enable_memory_compaction=enable_memory_compaction,
+            memory_manager=memory_manager,
+            memory_compact_threshold=memory_compact_threshold,
+        )
+
+        # Initialize team infrastructure if available
+        self._messenger: Optional["AgentMessenger"] = None
+        self._task_delegator: Optional["TaskDelegator"] = None
+        if TEAM_INFRA_AVAILABLE:
+            self._init_team_infrastructure()
+
+        logger.info(
+            "EvoAgent initialized: %s (workspace: %s)",
+            agent_id,
+            workspace_dir,
+        )
+
+    def _load_agent_config(self) -> Dict[str, Any]:
+        """Load agent configuration from workspace.
+
+        Returns:
+            Agent configuration dictionary
+        """
+        config_path = self.workspace_dir / "agent.yaml"
+        if config_path.exists():
+            loaded = load_agent_workspace_config(config_path)
+            return dict(loaded.values)
+        return {}
+
+    def _build_system_prompt(self) -> str:
+        """Build system prompt from workspace markdown files.
+
+        Uses PromptBuilder to load and combine AGENTS.md, SOUL.md,
+        PROFILE.md, and other configured files.
+
+        Returns:
+            Complete system prompt string
+        """
+        prompt = build_system_prompt_from_workspace(
+            workspace_dir=self.workspace_dir,
+            enabled_files=self._prompt_files,
+            agent_id=self.agent_id,
+            extra_context=self._env_context,
+        )
+        return prompt
+
+    def _create_toolkit(self) -> Toolkit:
+        """Create and populate toolkit with agent skills.
+
+        Loads skills from the agent's active skills directory and
+        registers them with the toolkit.
+
+        Returns:
+            Configured Toolkit instance
+        """
+        toolkit = Toolkit(
+            agent_skill_instruction=(
+                "<system-info>You have access to specialized skills. "
+                "Each skill lives in a directory and is described by SKILL.md. "
+                "Follow the skill instructions when they are relevant to the current task."
+                "</system-info>"
+            ),
+            agent_skill_template="- {name} (dir: {dir}): {description}",
+        )
+
+        # Register skills from active directory
+        active_skills_dir = self._skills_manager.get_agent_active_root(
+            self.config_name,
+            self.agent_id,
+        )
+
+        if active_skills_dir.exists():
+            for skill_dir in sorted(active_skills_dir.iterdir()):
+                if skill_dir.is_dir() and (skill_dir / "SKILL.md").exists():
+                    try:
+                        toolkit.register_agent_skill(str(skill_dir))
+                        logger.debug("Registered skill: %s", skill_dir.name)
+                    except Exception as e:
+                        logger.error(
+                            "Failed to register skill '%s': %s",
+                            skill_dir.name,
+                            e,
+                        )
+
+        return toolkit
+
+    def _register_hooks(
+        self,
+        enable_bootstrap: bool,
+        enable_memory_compaction: bool,
+        memory_manager: Optional[Any],
+        memory_compact_threshold: Optional[int],
+    ) -> None:
+        """Register agent hooks.
+
+        Args:
+            enable_bootstrap: Enable bootstrap hook
+            enable_memory_compaction: Enable memory compaction hook
+            memory_manager: Memory manager instance
+            memory_compact_threshold: Token threshold for compaction
+        """
+        # Bootstrap hook - checks BOOTSTRAP.md on first interaction
+        if enable_bootstrap:
+            bootstrap_hook = BootstrapHook(
+                workspace_dir=self.workspace_dir,
+                language="zh",
+            )
+            self._hook_manager.register(
+                hook_type=HOOK_PRE_REASONING,
+                hook_name="bootstrap",
+                hook=bootstrap_hook,
+            )
+            logger.debug("Registered bootstrap hook")
+
+        # Memory compaction hook
+        if enable_memory_compaction and memory_manager is not None:
+            compaction_hook = MemoryCompactionHook(
+                memory_manager=memory_manager,
+                memory_compact_threshold=memory_compact_threshold,
+            )
+            self._hook_manager.register(
+                hook_type=HOOK_PRE_REASONING,
+                hook_name="memory_compaction",
+                hook=compaction_hook,
+            )
+            logger.debug("Registered memory compaction hook")
+
+        # Workspace watch hook - auto-reload markdown files on change
+        workspace_watch_hook = WorkspaceWatchHook(
+            workspace_dir=self.workspace_dir,
+        )
+        self._hook_manager.register(
+            hook_type=HOOK_PRE_REASONING,
+            hook_name="workspace_watch",
+            hook=workspace_watch_hook,
+        )
+        logger.debug("Registered workspace watch hook")
+
+    async def _reasoning(self, **kwargs) -> Msg:
+        """Override reasoning to execute pre-reasoning hooks.
+
+        Args:
+            **kwargs: Arguments for reasoning
+
+        Returns:
+            Response message
+        """
+        # Execute pre-reasoning hooks
+        kwargs = await self._hook_manager.execute(
+            hook_type=HOOK_PRE_REASONING,
+            agent=self,
+            kwargs=kwargs,
+        )
+
+        # Call parent (which may be ToolGuardMixin's _reasoning)
+        return await super()._reasoning(**kwargs)
+
+    def reload_skills(self, active_skill_dirs: Optional[List[Path]] = None) -> None:
+        """Reload skills at runtime.
+
+        Rebuilds the toolkit with current skills from the active directory.
+
+        Args:
+            active_skill_dirs: Optional list of specific skill directories to load
+        """
+        logger.info("Reloading skills for agent: %s", self.agent_id)
+
+        # Create new toolkit
+        new_toolkit = Toolkit(
+            agent_skill_instruction=(
+                "<system-info>You have access to specialized skills. "
+                "Each skill lives in a directory and is described by SKILL.md. "
+                "Follow the skill instructions when they are relevant to the current task."
+                "</system-info>"
+            ),
+            agent_skill_template="- {name} (dir: {dir}): {description}",
+        )
+
+        # Register skills
+        if active_skill_dirs is None:
+            active_skills_dir = self._skills_manager.get_agent_active_root(
+                self.config_name,
+                self.agent_id,
+            )
+            if active_skills_dir.exists():
+                active_skill_dirs = [
+                    d for d in active_skills_dir.iterdir()
+                    if d.is_dir() and (d / "SKILL.md").exists()
+                ]
+            else:
+                active_skill_dirs = []
+
+        for skill_dir in active_skill_dirs:
+            if skill_dir.exists() and (skill_dir / "SKILL.md").exists():
+                try:
+                    new_toolkit.register_agent_skill(str(skill_dir))
+                    logger.debug("Reloaded skill: %s", skill_dir.name)
+                except Exception as e:
+                    logger.error(
+                        "Failed to reload skill '%s': %s",
+                        skill_dir.name,
+                        e,
+                    )
+
+        # Replace toolkit
+        self.toolkit = new_toolkit
+        logger.info("Skills reloaded for agent: %s", self.agent_id)
+
+    def rebuild_sys_prompt(self) -> None:
+        """Rebuild and replace the system prompt at runtime.
+
+        Useful after updating AGENTS.md, SOUL.md, PROFILE.md, etc.
+        to ensure the prompt reflects the latest configuration.
+
+        Updates both self._sys_prompt and the first system-role
+        message stored in self.memory.content.
+        """
+        logger.info("Rebuilding system prompt for agent: %s", self.agent_id)
+
+        # Reload agent config in case it changed
+        self._agent_config = self._load_agent_config()
+
+        # Rebuild prompt
+        self._sys_prompt = self._build_system_prompt()
+
+        # Update memory if system message exists
+        if hasattr(self, "memory") and self.memory.content:
+            for msg, _marks in self.memory.content:
+                if getattr(msg, "role", None) == "system":
+                    msg.content = self._sys_prompt
+                    logger.debug("Updated system message in memory")
+                    break
+
+        logger.info("System prompt rebuilt for agent: %s", self.agent_id)
+
+    async def reply(
+        self,
+        msg: Msg | List[Msg] | None = None,
+        structured_model: Optional[Type[Any]] = None,
+    ) -> Msg:
+        """Process a message and return a response.
+
+        Args:
+            msg: Input message(s) from user
+            structured_model: Optional pydantic model for structured output
+
+        Returns:
+            Response message
+        """
+        # Handle list of messages
+        if isinstance(msg, list):
+            # Process each message in sequence
+            for m in msg[:-1]:
+                await self.memory.add(m)
+            msg = msg[-1] if msg else None
+
+        return await super().reply(msg=msg, structured_model=structured_model)
+
+    def get_agent_info(self) -> Dict[str, Any]:
+        """Get agent information.
+
+        Returns:
+            Dictionary with agent metadata
+        """
+        return {
+            "agent_id": self.agent_id,
+            "config_name": self.config_name,
+            "workspace_dir": str(self.workspace_dir),
+            "skills_count": len([
+                s for s in self._skills_manager.list_active_skill_metadata(
+                    self.config_name,
+                    self.agent_id,
+                )
+            ]),
+            "registered_hooks": self._hook_manager.list_hooks(),
+            "team_infra_available": TEAM_INFRA_AVAILABLE,
+        }
+
+    def _init_team_infrastructure(self) -> None:
+        """Initialize team infrastructure components (messenger and task delegator).
+
+        This method initializes the AgentMessenger for inter-agent communication
+        and the TaskDelegator for subagent delegation.
+        """
+        if not TEAM_INFRA_AVAILABLE:
+            return
+
+        try:
+            self._messenger = AgentMessenger(agent_id=self.agent_id)
+            self._task_delegator = TaskDelegator(agent=self)
+            logger.debug(
+                "Team infrastructure initialized for agent: %s",
+                self.agent_id,
+            )
+        except Exception as e:
+            logger.warning(
+                "Failed to initialize team infrastructure for %s: %s",
+                self.agent_id,
+                e,
+            )
+            self._messenger = None
+            self._task_delegator = None
+
+    @property
+    def messenger(self) -> Optional["AgentMessenger"]:
+        """Get the agent's messenger for inter-agent communication.
+
+        Returns:
+            AgentMessenger instance if available, None otherwise
+        """
+        return self._messenger
+
+    async def delegate_task(
+        self,
+        task_type: str,
+        task_data: Dict[str, Any],
+        target_agent: Optional[str] = None,
+    ) -> Dict[str, Any]:
+        """Delegate a task to a subagent using the TaskDelegator.
+
+        Args:
+            task_type: Type of task to delegate
+            task_data: Data/payload for the task
+            target_agent: Optional specific agent ID to delegate to
+
+        Returns:
+            Dict containing the delegation result
+        """
+        if not TEAM_INFRA_AVAILABLE or self._task_delegator is None:
+            return {
+                "success": False,
+                "error": "Team infrastructure not available",
+            }
+
+        try:
+            return await self._task_delegator.delegate_task(
+                task_type=task_type,
+                task_data=task_data,
+                target_agent=target_agent,
+            )
+        except Exception as e:
+            logger.error(
+                "Task delegation failed for %s: %s",
+                self.agent_id,
+                e,
+            )
+            return {"success": False, "error": str(e)}
+
+
+__all__ = ["EvoAgent"]
--- a/backend/agents/base/hooks.py
+++ b/backend/agents/base/hooks.py
@@ -0,0 +1,702 @@
+# -*- coding: utf-8 -*-
+"""Hook system for EvoAgent.
+
+Provides pre_reasoning and post_acting hooks with built-in implementations:
+- BootstrapHook: First-time setup guidance
+- MemoryCompactionHook: Automatic memory compression
+
+Based on CoPaw's hooks design.
+"""
+from __future__ import annotations
+
+import logging
+from abc import ABC, abstractmethod
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from agentscope.agent import ReActAgent
+
+logger = logging.getLogger(__name__)
+
+# Hook types
+HookType = str
+HOOK_PRE_REASONING: HookType = "pre_reasoning"
+HOOK_POST_ACTING: HookType = "post_acting"
+
+
+class Hook(ABC):
+    """Abstract base class for agent hooks."""
+
+    @abstractmethod
+    async def __call__(
+        self,
+        agent: "ReActAgent",
+        kwargs: Dict[str, Any],
+    ) -> Optional[Dict[str, Any]]:
+        """Execute the hook.
+
+        Args:
+            agent: The agent instance
+            kwargs: Input arguments to the method being hooked
+
+        Returns:
+            Modified kwargs or None to use original
+        """
+        pass
+
+
+class HookManager:
+    """Manages agent hooks.
+
+    Provides registration and execution of hooks for different
+    lifecycle events in the agent's operation.
+    """
+
+    def __init__(self):
+        self._hooks: Dict[HookType, List[tuple[str, Hook]]] = {
+            HOOK_PRE_REASONING: [],
+            HOOK_POST_ACTING: [],
+        }
+
+    def register(
+        self,
+        hook_type: HookType,
+        hook_name: str,
+        hook: Hook | Callable,
+    ) -> None:
+        """Register a hook.
+
+        Args:
+            hook_type: Type of hook (pre_reasoning, post_acting)
+            hook_name: Unique name for this hook
+            hook: Hook instance or callable
+        """
+        # Remove existing hook with same name
+        self._hooks[hook_type] = [
+            (name, h) for name, h in self._hooks[hook_type] if name != hook_name
+        ]
+        self._hooks[hook_type].append((hook_name, hook))
+        logger.debug("Registered hook '%s' for type '%s'", hook_name, hook_type)
+
+    def unregister(self, hook_type: HookType, hook_name: str) -> bool:
+        """Unregister a hook.
+
+        Args:
+            hook_type: Type of hook
+            hook_name: Name of the hook to remove
+
+        Returns:
+            True if hook was found and removed
+        """
+        original_len = len(self._hooks[hook_type])
+        self._hooks[hook_type] = [
+            (name, h) for name, h in self._hooks[hook_type] if name != hook_name
+        ]
+        removed = len(self._hooks[hook_type]) < original_len
+        if removed:
+            logger.debug("Unregistered hook '%s' from type '%s'", hook_name, hook_type)
+        return removed
+
+    async def execute(
+        self,
+        hook_type: HookType,
+        agent: "ReActAgent",
+        kwargs: Dict[str, Any],
+    ) -> Dict[str, Any]:
+        """Execute all hooks of a given type.
+
+        Args:
+            hook_type: Type of hooks to execute
+            agent: The agent instance
+            kwargs: Input arguments
+
+        Returns:
+            Potentially modified kwargs
+        """
+        for name, hook in self._hooks[hook_type]:
+            try:
+                result = await hook(agent, kwargs)
+                if result is not None:
+                    kwargs = result
+            except Exception as e:
+                logger.error("Hook '%s' failed: %s", name, e, exc_info=True)
+
+        return kwargs
+
+    def list_hooks(self, hook_type: Optional[HookType] = None) -> List[str]:
+        """List registered hook names.
+
+        Args:
+            hook_type: Optional type to filter by
+
+        Returns:
+            List of hook names
+        """
+        if hook_type:
+            return [name for name, _ in self._hooks.get(hook_type, [])]
+
+        names = []
+        for hooks in self._hooks.values():
+            names.extend([name for name, _ in hooks])
+        return names
+
+
+class BootstrapHook(Hook):
+    """Hook for bootstrap guidance on first user interaction.
+
+    This hook looks for a BOOTSTRAP.md file in the working directory
+    and if found, prepends guidance to the first user message to help
+    establish the agent's identity and user preferences.
+    """
+
+    def __init__(
+        self,
+        workspace_dir: Path,
+        language: str = "zh",
+    ):
+        """Initialize bootstrap hook.
+
+        Args:
+            workspace_dir: Working directory containing BOOTSTRAP.md
+            language: Language code for bootstrap guidance (en/zh)
+        """
+        self.workspace_dir = Path(workspace_dir)
+        self.language = language
+        self._completed_flag = self.workspace_dir / ".bootstrap_completed"
+
+    def _is_first_user_interaction(self, agent: "ReActAgent") -> bool:
+        """Check if this is the first user interaction.
+
+        Args:
+            agent: The agent instance
+
+        Returns:
+            True if first user interaction
+        """
+        if not hasattr(agent, "memory") or not agent.memory.content:
+            return True
+
+        # Count user messages (excluding system)
+        user_count = sum(
+            1 for msg, _ in agent.memory.content if msg.role == "user"
+        )
+        return user_count <= 1
+
+    def _build_bootstrap_guidance(self) -> str:
+        """Build bootstrap guidance message.
+
+        Returns:
+            Formatted bootstrap guidance
+        """
+        if self.language == "zh":
+            return (
+                "# 引导模式\n"
+                "\n"
+                "工作目录中存在 `BOOTSTRAP.md` — 首次设置。\n"
+                "\n"
+                "1. 阅读 BOOTSTRAP.md，友好地表示初次见面，"
+                "引导用户完成设置。\n"
+                "2. 按照 BOOTSTRAP.md 的指示，"
+                "帮助用户定义你的身份和偏好。\n"
+                "3. 按指南创建/更新必要文件"
+                "（PROFILE.md、MEMORY.md 等）。\n"
+                "4. 完成后删除 BOOTSTRAP.md。\n"
+                "\n"
+                "如果用户希望跳过，直接回答下面的问题即可。\n"
+                "\n"
+                "---\n"
+                "\n"
+            )
+
+        return (
+            "# BOOTSTRAP MODE\n"
+            "\n"
+            "`BOOTSTRAP.md` exists — first-time setup.\n"
+            "\n"
+            "1. Read BOOTSTRAP.md, greet the user, "
+            "and guide them through setup.\n"
+            "2. Follow BOOTSTRAP.md instructions "
+            "to define identity and preferences.\n"
+            "3. Create/update files "
+            "(PROFILE.md, MEMORY.md, etc.) as described.\n"
+            "4. Delete BOOTSTRAP.md when done.\n"
+            "\n"
+            "If the user wants to skip, answer their "
+            "question directly instead.\n"
+            "\n"
+            "---\n"
+            "\n"
+        )
+
+    async def __call__(
+        self,
+        agent: "ReActAgent",
+        kwargs: Dict[str, Any],
+    ) -> Optional[Dict[str, Any]]:
+        """Check and load BOOTSTRAP.md on first user interaction.
+
+        Args:
+            agent: The agent instance
+            kwargs: Input arguments to the _reasoning method
+
+        Returns:
+            None (hook doesn't modify kwargs)
+        """
+        try:
+            bootstrap_path = self.workspace_dir / "BOOTSTRAP.md"
+
+            # Check if bootstrap has already been triggered
+            if self._completed_flag.exists():
+                return None
+
+            if not bootstrap_path.exists():
+                return None
+
+            if not self._is_first_user_interaction(agent):
+                return None
+
+            bootstrap_guidance = self._build_bootstrap_guidance()
+
+            logger.debug("Found BOOTSTRAP.md [%s], prepending guidance", self.language)
+
+            # Prepend to first user message in memory
+            if hasattr(agent, "memory") and agent.memory.content:
+                system_count = sum(
+                    1 for msg, _ in agent.memory.content if msg.role == "system"
+                )
+                for msg, _ in agent.memory.content[system_count:]:
+                    if msg.role == "user":
+                        # Prepend guidance to message content
+                        original_content = msg.content
+                        msg.content = bootstrap_guidance + original_content
+                        break
+
+            logger.debug("Bootstrap guidance prepended to first user message")
+
+            # Create completion flag to prevent repeated triggering
+            self._completed_flag.touch()
+            logger.debug("Created bootstrap completion flag")
+
+        except Exception as e:
+            logger.error("Failed to process bootstrap: %s", e, exc_info=True)
+
+        return None
+
+
+class WorkspaceWatchHook(Hook):
+    """Hook for auto-reloading workspace markdown files on change.
+
+    Monitors SOUL.md, AGENTS.md, PROFILE.md, etc. and triggers
+    a prompt rebuild when any of them change. Based on CoPaw's
+    AgentConfigWatcher approach but for markdown files.
+    """
+
+    # Files to monitor (same as PromptBuilder.DEFAULT_FILES)
+    WATCHED_FILES = frozenset([
+        "SOUL.md", "AGENTS.md", "PROFILE.md", "ROLE.md",
+        "POLICY.md", "MEMORY.md", "HEARTBEAT.md", "STYLE.md",
+        "BOOTSTRAP.md",
+    ])
+
+    def __init__(
+        self,
+        workspace_dir: Path,
+        poll_interval: float = 2.0,
+    ):
+        """Initialize workspace watch hook.
+
+        Args:
+            workspace_dir: Workspace directory to monitor
+            poll_interval: How often to check for changes (seconds)
+        """
+        self.workspace_dir = Path(workspace_dir)
+        self.poll_interval = poll_interval
+        self._last_mtimes: dict[str, float] = {}
+        self._initialized = False
+
+    def _scan_mtimes(self) -> dict[str, float]:
+        """Scan watched files and return their current mtimes."""
+        mtimes = {}
+        for name in self.WATCHED_FILES:
+            path = self.workspace_dir / name
+            if path.exists():
+                mtimes[name] = path.stat().st_mtime
+        return mtimes
+
+    def _has_changes(self) -> bool:
+        """Check if any watched file has changed since last check."""
+        current = self._scan_mtimes()
+
+        if not self._initialized:
+            self._last_mtimes = current
+            self._initialized = True
+            return False
+
+        # Check for new, modified, or deleted files
+        if set(current.keys()) != set(self._last_mtimes.keys()):
+            self._last_mtimes = current
+            return True
+
+        for name, mtime in current.items():
+            if mtime != self._last_mtimes.get(name):
+                self._last_mtimes = current
+                return True
+
+        return False
+
+    async def __call__(
+        self,
+        agent: "ReActAgent",
+        kwargs: Dict[str, Any],
+    ) -> Optional[Dict[str, Any]]:
+        """Check for file changes and rebuild prompt if needed.
+
+        Args:
+            agent: The agent instance
+            kwargs: Input arguments (unused)
+
+        Returns:
+            None
+        """
+        try:
+            if self._has_changes():
+                logger.info(
+                    "Workspace files changed, triggering prompt rebuild for: %s",
+                    getattr(agent, "agent_id", "unknown"),
+                )
+                if hasattr(agent, "rebuild_sys_prompt"):
+                    agent.rebuild_sys_prompt()
+                else:
+                    logger.warning(
+                        "Agent %s has no rebuild_sys_prompt method",
+                        getattr(agent, "agent_id", "unknown"),
+                    )
+        except Exception as e:
+            logger.error("Workspace watch hook failed: %s", e, exc_info=True)
+
+        return None
+
+
+class MemoryCompactionHook(Hook):
+    """Hook for automatic memory compaction when context is full.
+
+    This hook monitors the token count of messages and triggers compaction
+    when it exceeds the threshold. It preserves the system prompt and recent
+    messages while summarizing older conversation history.
+
+    Based on CoPaw's memory compaction design with additional improvements:
+    - memory_compact_ratio: Ratio to compact when threshold reached
+    - memory_reserve_ratio: Always keep a reserve of tokens for recent messages
+    - enable_tool_result_compact: Compact tool results separately
+    - tool_result_compact_keep_n: Number of tool results to keep
+    """
+
+    def __init__(
+        self,
+        memory_manager: Any,
+        memory_compact_threshold: Optional[int] = None,
+        memory_compact_ratio: float = 0.75,
+        memory_reserve_ratio: float = 0.1,
+        enable_tool_result_compact: bool = False,
+        tool_result_compact_keep_n: int = 5,
+    ):
+        """Initialize memory compaction hook.
+
+        Args:
+            memory_manager: Memory manager instance for compaction
+            memory_compact_threshold: Token threshold for compaction
+            memory_compact_ratio: Target ratio to compact to (e.g., 0.75 = compact to 75%)
+            memory_reserve_ratio: Reserve ratio to always keep free (e.g., 0.1 = 10%)
+            enable_tool_result_compact: Enable tool result compaction
+            tool_result_compact_keep_n: Number of tool results to keep
+        """
+        self.memory_manager = memory_manager
+        self.memory_compact_threshold = memory_compact_threshold
+        self.memory_compact_ratio = memory_compact_ratio
+        self.memory_reserve_ratio = memory_reserve_ratio
+        self.enable_tool_result_compact = enable_tool_result_compact
+        self.tool_result_compact_keep_n = tool_result_compact_keep_n
+
+    async def __call__(
+        self,
+        agent: "ReActAgent",
+        kwargs: Dict[str, Any],
+    ) -> Optional[Dict[str, Any]]:
+        """Pre-reasoning hook to check and compact memory if needed.
+
+        Args:
+            agent: The agent instance
+            kwargs: Input arguments to the _reasoning method
+
+        Returns:
+            None (hook doesn't modify kwargs)
+        """
+        try:
+            if not hasattr(agent, "memory") or not self.memory_manager:
+                return None
+
+            memory = agent.memory
+
+            # Get current token count estimate
+            messages = await memory.get_memory()
+            total_tokens = self._estimate_tokens(messages)
+
+            if self.memory_compact_threshold is None:
+                return None
+
+            if total_tokens < self.memory_compact_threshold:
+                return None
+
+            logger.info(
+                "Memory compaction triggered: %d tokens (threshold: %d)",
+                total_tokens,
+                self.memory_compact_threshold,
+            )
+
+            # Compact memory
+            await self._compact_memory(agent, messages)
+
+        except Exception as e:
+            logger.error("Failed to compact memory: %s", e, exc_info=True)
+
+        return None
+
+    def _estimate_tokens(self, messages: List[Any]) -> int:
+        """Estimate token count for messages.
+
+        Args:
+            messages: List of messages
+
+        Returns:
+            Estimated token count
+        """
+        # Simple estimation: ~4 chars per token
+        total_chars = sum(
+            len(str(getattr(msg, "content", "")))
+            for msg in messages
+        )
+        return total_chars // 4
+
+    async def _compact_memory(
+        self,
+        agent: "ReActAgent",
+        messages: List[Any],
+    ) -> None:
+        """Compact memory by summarizing older messages.
+
+        Uses CoPaw-style memory management:
+        - memory_compact_ratio: Target ratio to compact to (e.g., 0.75 means compact to 75%)
+        - memory_reserve_ratio: Always keep this ratio free (e.g., 0.1 means keep 10% for recent)
+
+        Args:
+            agent: The agent instance
+            messages: Current messages in memory
+        """
+        if self.memory_compact_threshold is None:
+            return
+
+        # Estimate total tokens
+        total_tokens = self._estimate_tokens(messages)
+
+        # Calculate reserve based on ratio (CoPaw-style)
+        reserve_tokens = int(total_tokens * self.memory_reserve_ratio)
+
+        # Calculate target tokens after compaction
+        target_tokens = int(total_tokens * self.memory_compact_ratio)
+        target_tokens = max(target_tokens, total_tokens - reserve_tokens)
+
+        # Find messages to compact (older ones)
+        # Keep recent messages that fit within target
+        messages_to_compact = []
+        kept_tokens = 0
+
+        # Start from oldest, stop when we've kept enough
+        for msg in messages:
+            msg_tokens = self._estimate_tokens([msg])
+            if kept_tokens + msg_tokens > target_tokens:
+                messages_to_compact.append(msg)
+            else:
+                kept_tokens += msg_tokens
+
+        if not messages_to_compact:
+            return
+
+        logger.info(
+            "Compacting %d messages (%d tokens) to target %d tokens",
+            len(messages_to_compact),
+            self._estimate_tokens(messages_to_compact),
+            target_tokens,
+        )
+
+        # Use memory manager to compact if available
+        if hasattr(self.memory_manager, "compact_memory"):
+            try:
+                summary = await self.memory_manager.compact_memory(
+                    messages=messages_to_compact,
+                )
+                logger.info(
+                    "Memory compacted: %d messages summarized, summary: %s",
+                    len(messages_to_compact),
+                    summary[:200] if summary else "N/A",
+                )
+
+                # Mark messages as compressed if supported
+                if hasattr(agent.memory, "update_messages_mark"):
+                    from agentscope.agent._react_agent import _MemoryMark
+                    await agent.memory.update_messages_mark(
+                        new_mark=_MemoryMark.COMPRESSED,
+                        msg_ids=[msg.id for msg in messages_to_compact],
+                    )
+
+            except Exception as e:
+                logger.error("Memory manager compaction failed: %s", e)
+
+        # Tool result compaction (CoPaw-style)
+        if self.enable_tool_result_compact:
+            await self._compact_tool_results(agent, messages)
+
+    async def _compact_tool_results(
+        self,
+        agent: "ReActAgent",
+        messages: List[Any],
+    ) -> None:
+        """Compact tool results by keeping only recent ones.
+
+        Based on CoPaw's tool_result_compact_keep_n pattern.
+        Tool results can be very verbose, so we keep only the N most recent ones.
+
+        Args:
+            agent: The agent instance
+            messages: Current messages in memory
+        """
+        if not hasattr(agent.memory, "content"):
+            return
+
+        # Find tool result messages (usually have "tool" role or tool_related content)
+        tool_results = []
+        for msg, _ in agent.memory.content:
+            if hasattr(msg, "role") and msg.role == "tool":
+                tool_results.append(msg)
+
+        if len(tool_results) <= self.tool_result_compact_keep_n:
+            return
+
+        # Keep only the most recent N tool results
+        excess_results = tool_results[:-self.tool_result_compact_keep_n]
+
+        logger.info(
+            "Tool result compaction: %d tool results found, keeping %d, compacting %d",
+            len(tool_results),
+            self.tool_result_compact_keep_n,
+            len(excess_results),
+        )
+
+        # Mark excess tool results as compressed if supported
+        if hasattr(agent.memory, "update_messages_mark"):
+            from agentscope.agent._react_agent import _MemoryMark
+            await agent.memory.update_messages_mark(
+                new_mark=_MemoryMark.COMPRESSED,
+                msg_ids=[msg.id for msg in excess_results],
+            )
+
+
+class HeartbeatHook(Hook):
+    """Pre-reasoning hook that injects HEARTBEAT.md content.
+
+    Reads the agent's HEARTBEAT.md file and prepends it to the
+    reasoning input, causing the agent to perform self-checks.
+
+    This enables "主动检查" (proactive monitoring) - periodic
+    market condition and position checks during trading hours.
+    """
+
+    HEARTBEAT_FILE = "HEARTBEAT.md"
+
+    def __init__(self, workspace_dir: Path):
+        """Initialize heartbeat hook.
+
+        Args:
+            workspace_dir: Working directory containing HEARTBEAT.md
+        """
+        self.workspace_dir = Path(workspace_dir)
+        self._completed_flag = self.workspace_dir / ".heartbeat_completed"
+
+    def _read_heartbeat_content(self) -> Optional[str]:
+        """Read HEARTBEAT.md if it exists and is non-empty.
+
+        Returns:
+            The HEARTBEAT.md content stripped of whitespace, or None
+            if the file is absent or empty.
+        """
+        hb_path = self.workspace_dir / self.HEARTBEAT_FILE
+        if not hb_path.exists():
+            return None
+        content = hb_path.read_text(encoding="utf-8").strip()
+        return content if content else None
+
+    async def __call__(
+        self,
+        agent: "ReActAgent",
+        kwargs: Dict[str, Any],
+    ) -> Optional[Dict[str, Any]]:
+        """Prepend heartbeat task to user message.
+
+        Args:
+            agent: The agent instance
+            kwargs: Input arguments to the _reasoning method
+
+        Returns:
+            Modified kwargs with heartbeat content prepended, or None
+            if no HEARTBEAT.md content is available.
+        """
+        try:
+            content = self._read_heartbeat_content()
+            if not content:
+                return None
+
+            logger.debug(
+                "Heartbeat: found HEARTBEAT.md for agent %s",
+                getattr(agent, "agent_id", "unknown"),
+            )
+
+            # Build heartbeat task instruction (Chinese)
+            hb_task = (
+                "# 定期主动检查\n\n"
+                f"{content}\n\n"
+                "请执行上述检查并报告结果。"
+            )
+
+            # Inject into the first user message in memory
+            if hasattr(agent, "memory") and agent.memory.content:
+                system_count = sum(
+                    1 for msg, _ in agent.memory.content if msg.role == "system"
+                )
+                for msg, _ in agent.memory.content[system_count:]:
+                    if msg.role == "user":
+                        original_content = msg.content
+                        msg.content = hb_task + "\n\n" + original_content
+                        break
+
+            logger.debug(
+                "Heartbeat task prepended for agent %s",
+                getattr(agent, "agent_id", "unknown"),
+            )
+
+        except Exception as e:
+            logger.error("Heartbeat hook failed: %s", e, exc_info=True)
+
+        return None
+
+
+__all__ = [
+    "Hook",
+    "HookManager",
+    "HookType",
+    "HOOK_PRE_REASONING",
+    "HOOK_POST_ACTING",
+    "BootstrapHook",
+    "HeartbeatHook",
+    "MemoryCompactionHook",
+    "WorkspaceWatchHook",
+]
--- a/backend/agents/base/skill_adaptation_hook.py
+++ b/backend/agents/base/skill_adaptation_hook.py
@@ -0,0 +1,489 @@
+# -*- coding: utf-8 -*-
+"""Skill adaptation hook for automatic evaluation-to-iteration闭环.
+
+Monitors evaluation metrics against configurable thresholds and triggers
+automatic skill reload or logs warnings when thresholds are breached.
+"""
+from __future__ import annotations
+
+import json
+import logging
+from dataclasses import dataclass, field
+from datetime import datetime
+from enum import Enum
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Set
+
+from .evaluation_hook import (
+    EvaluationCollector,
+    EvaluationResult,
+    MetricType,
+)
+
+logger = logging.getLogger(__name__)
+
+
+class AdaptationAction(Enum):
+    """Actions to take when threshold is breached."""
+    RELOAD = "reload"           # 自动重新加载技能
+    WARN = "warn"               # 记录警告供人工审核
+    BOTH = "both"               # 同时执行重载和警告
+    NONE = "none"               # 不做任何操作
+
+
+@dataclass
+class AdaptationThreshold:
+    """Threshold configuration for a metric."""
+    metric_type: MetricType
+    operator: str = "lt"        # lt (less than), gt (greater than), lte, gte, eq
+    value: float = 0.0
+    window_size: int = 10       # 移动窗口大小，用于计算滑动平均
+    min_samples: int = 5        # 最少样本数才触发检查
+    action: AdaptationAction = AdaptationAction.WARN
+    cooldown_seconds: int = 300  # 触发后的冷却时间
+
+    def evaluate(self, current_value: float) -> bool:
+        """Evaluate if threshold is breached."""
+        ops = {
+            "lt": lambda x, y: x < y,
+            "lte": lambda x, y: x <= y,
+            "gt": lambda x, y: x > y,
+            "gte": lambda x, y: x >= y,
+            "eq": lambda x, y: x == y,
+        }
+        op_func = ops.get(self.operator)
+        if op_func is None:
+            logger.warning(f"Unknown operator: {self.operator}")
+            return False
+        return op_func(current_value, self.value)
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "metric_type": self.metric_type.value,
+            "operator": self.operator,
+            "value": self.value,
+            "window_size": self.window_size,
+            "min_samples": self.min_samples,
+            "action": self.action.value,
+            "cooldown_seconds": self.cooldown_seconds,
+        }
+
+
+@dataclass
+class AdaptationEvent:
+    """Record of an adaptation trigger event."""
+    timestamp: str
+    skill_name: str
+    metric_type: MetricType
+    threshold: AdaptationThreshold
+    current_value: float
+    avg_value: float
+    action_taken: AdaptationAction
+    details: Dict[str, Any] = field(default_factory=dict)
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "timestamp": self.timestamp,
+            "skill_name": self.skill_name,
+            "metric_type": self.metric_type.value,
+            "threshold": self.threshold.to_dict(),
+            "current_value": self.current_value,
+            "avg_value": self.avg_value,
+            "action_taken": self.action_taken.value,
+            "details": self.details,
+        }
+
+
+class SkillAdaptationHook:
+    """Hook for monitoring evaluation metrics and triggering skill adaptation.
+
+    This hook wraps EvaluationHook to add threshold-based adaptation logic.
+    When metrics breach configured thresholds, it can:
+    - Automatically reload skills via SkillsManager
+    - Log warnings for human review
+    - Both
+    """
+
+    # Default thresholds for common metrics
+    DEFAULT_THRESHOLDS: List[AdaptationThreshold] = [
+        AdaptationThreshold(
+            metric_type=MetricType.HIT_RATE,
+            operator="lt",
+            value=0.5,
+            action=AdaptationAction.WARN,
+            cooldown_seconds=600,
+        ),
+        AdaptationThreshold(
+            metric_type=MetricType.RISK_VIOLATION,
+            operator="gt",
+            value=0.1,
+            action=AdaptationAction.WARN,
+            cooldown_seconds=300,
+        ),
+        AdaptationThreshold(
+            metric_type=MetricType.DECISION_LATENCY,
+            operator="gt",
+            value=5000,  # 5 seconds
+            action=AdaptationAction.WARN,
+            cooldown_seconds=300,
+        ),
+    ]
+
+    def __init__(
+        self,
+        storage_dir: Path,
+        run_id: str,
+        agent_id: str,
+        thresholds: Optional[List[AdaptationThreshold]] = None,
+        collector: Optional[EvaluationCollector] = None,
+    ):
+        """Initialize skill adaptation hook.
+
+        Args:
+            storage_dir: Directory to store adaptation events
+            run_id: Current run identifier
+            agent_id: Current agent identifier
+            thresholds: Custom threshold configurations (uses defaults if None)
+            collector: Optional EvaluationCollector for historical data
+        """
+        self.storage_dir = Path(storage_dir)
+        self.run_id = run_id
+        self.agent_id = agent_id
+        self.thresholds = thresholds or self.DEFAULT_THRESHOLDS
+        self.collector = collector or EvaluationCollector(storage_dir)
+
+        # Track cooldowns to prevent rapid re-triggering
+        self._cooldowns: Dict[str, datetime] = {}
+
+        # Store recent metrics in memory for quick access
+        self._recent_metrics: Dict[str, List[float]] = {}
+
+        # Pending adaptation events
+        self._pending_events: List[AdaptationEvent] = []
+
+    def check_threshold(
+        self,
+        skill_name: str,
+        metric_type: MetricType,
+        current_value: float,
+    ) -> Optional[AdaptationEvent]:
+        """Check if a metric breaches any threshold.
+
+        Args:
+            skill_name: Name of the skill
+            metric_type: Type of metric
+            current_value: Current metric value
+
+        Returns:
+            AdaptationEvent if threshold breached, None otherwise
+        """
+        # Find applicable thresholds
+        applicable_thresholds = [
+            t for t in self.thresholds
+            if t.metric_type == metric_type
+        ]
+
+        if not applicable_thresholds:
+            return None
+
+        # Check cooldown
+        cooldown_key = f"{skill_name}:{metric_type.value}"
+        now = datetime.now()
+        last_trigger = self._cooldowns.get(cooldown_key)
+
+        # Store current value first for avg calculation
+        self._store_metric(cooldown_key, current_value)
+
+        for threshold in applicable_thresholds:
+            if last_trigger:
+                elapsed = (now - last_trigger).total_seconds()
+                if elapsed < threshold.cooldown_seconds:
+                    continue
+
+            # Evaluate threshold
+            if threshold.evaluate(current_value):
+                # Calculate moving average
+                avg_value = self._calculate_avg(skill_name, metric_type, current_value)
+
+                # Check minimum samples (allow immediate trigger if min_samples <= 1)
+                sample_count = len(self._recent_metrics.get(cooldown_key, []))
+                if threshold.min_samples > 1 and sample_count < threshold.min_samples:
+                    # Not enough samples yet
+                    continue
+
+                # Trigger adaptation
+                event = AdaptationEvent(
+                    timestamp=now.isoformat(),
+                    skill_name=skill_name,
+                    metric_type=metric_type,
+                    threshold=threshold,
+                    current_value=current_value,
+                    avg_value=avg_value,
+                    action_taken=threshold.action,
+                    details={
+                        "run_id": self.run_id,
+                        "agent_id": self.agent_id,
+                    },
+                )
+
+                # Update cooldown
+                self._cooldowns[cooldown_key] = now
+
+                # Persist event
+                self._persist_event(event)
+
+                logger.info(
+                    f"Threshold breached for {skill_name}.{metric_type.value}: "
+                    f"current={current_value}, avg={avg_value}, action={threshold.action.value}"
+                )
+
+                return event
+
+        return None
+
+    def _calculate_avg(
+        self,
+        skill_name: str,
+        metric_type: MetricType,
+        current_value: float,
+    ) -> float:
+        """Calculate moving average for a metric."""
+        key = f"{skill_name}:{metric_type.value}"
+        values = self._recent_metrics.get(key, [])
+        if not values:
+            return current_value
+        return sum(values) / len(values)
+
+    def _store_metric(self, key: str, value: float) -> None:
+        """Store metric value with sliding window."""
+        if key not in self._recent_metrics:
+            self._recent_metrics[key] = []
+        self._recent_metrics[key].append(value)
+        # Keep only last 100 values
+        if len(self._recent_metrics[key]) > 100:
+            self._recent_metrics[key] = self._recent_metrics[key][-100:]
+
+    def _persist_event(self, event: AdaptationEvent) -> None:
+        """Persist adaptation event to storage."""
+        run_dir = self.storage_dir / self.run_id / "adaptations"
+        run_dir.mkdir(parents=True, exist_ok=True)
+
+        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
+        filename = f"{event.skill_name}_{event.metric_type.value}_{timestamp}.json"
+        filepath = run_dir / filename
+
+        try:
+            with open(filepath, "w", encoding="utf-8") as f:
+                json.dump(event.to_dict(), f, ensure_ascii=False, indent=2)
+            logger.debug(f"Persisted adaptation event to: {filepath}")
+        except Exception as e:
+            logger.error(f"Failed to persist adaptation event: {e}")
+
+        # Also add to pending list
+        self._pending_events.append(event)
+
+    def get_pending_warnings(self) -> List[AdaptationEvent]:
+        """Get all pending warning events that need human review."""
+        return [
+            e for e in self._pending_events
+            if e.action_taken in (AdaptationAction.WARN, AdaptationAction.BOTH)
+        ]
+
+    def clear_pending_warnings(self) -> None:
+        """Clear pending warnings after they have been reviewed."""
+        self._pending_events = [
+            e for e in self._pending_events
+            if e.action_taken == AdaptationAction.RELOAD
+        ]
+
+    def get_recent_events(
+        self,
+        skill_name: Optional[str] = None,
+        metric_type: Optional[MetricType] = None,
+        limit: int = 50,
+    ) -> List[AdaptationEvent]:
+        """Get recent adaptation events.
+
+        Args:
+            skill_name: Optional filter by skill name
+            metric_type: Optional filter by metric type
+            limit: Maximum number of events to return
+
+        Returns:
+            List of recent adaptation events
+        """
+        events_dir = self.storage_dir / self.run_id / "adaptations"
+        if not events_dir.exists():
+            return []
+
+        events = []
+        for eval_file in sorted(events_dir.glob("*.json"), reverse=True)[:limit]:
+            try:
+                with open(eval_file, "r", encoding="utf-8") as f:
+                    data = json.load(f)
+                    event = self._parse_event(data)
+                    if skill_name and event.skill_name != skill_name:
+                        continue
+                    if metric_type and event.metric_type != metric_type:
+                        continue
+                    events.append(event)
+            except Exception as e:
+                logger.warning(f"Failed to load adaptation event {eval_file}: {e}")
+
+        return events
+
+    def _parse_event(self, data: Dict[str, Any]) -> AdaptationEvent:
+        """Parse adaptation event from JSON data."""
+        threshold_data = data.get("threshold", {})
+        metric_type = MetricType(threshold_data.get("metric_type", "custom"))
+
+        threshold = AdaptationThreshold(
+            metric_type=metric_type,
+            operator=threshold_data.get("operator", "lt"),
+            value=threshold_data.get("value", 0.0),
+            window_size=threshold_data.get("window_size", 10),
+            min_samples=threshold_data.get("min_samples", 5),
+            action=AdaptationAction(threshold_data.get("action", "warn")),
+            cooldown_seconds=threshold_data.get("cooldown_seconds", 300),
+        )
+
+        return AdaptationEvent(
+            timestamp=data.get("timestamp", ""),
+            skill_name=data.get("skill_name", ""),
+            metric_type=metric_type,
+            threshold=threshold,
+            current_value=data.get("current_value", 0.0),
+            avg_value=data.get("avg_value", 0.0),
+            action_taken=AdaptationAction(data.get("action_taken", "warn")),
+            details=data.get("details", {}),
+        )
+
+    def add_threshold(self, threshold: AdaptationThreshold) -> None:
+        """Add a new threshold configuration."""
+        self.thresholds.append(threshold)
+
+    def remove_threshold(self, metric_type: MetricType) -> None:
+        """Remove all thresholds for a specific metric type."""
+        self.thresholds = [
+            t for t in self.thresholds
+            if t.metric_type != metric_type
+        ]
+
+    def update_threshold(
+        self,
+        metric_type: MetricType,
+        **kwargs,
+    ) -> None:
+        """Update threshold configuration for a metric type."""
+        for threshold in self.thresholds:
+            if threshold.metric_type == metric_type:
+                for key, value in kwargs.items():
+                    if hasattr(threshold, key):
+                        setattr(threshold, key, value)
+
+    def get_thresholds(self) -> List[AdaptationThreshold]:
+        """Get current threshold configurations."""
+        return list(self.thresholds)
+
+    def is_in_cooldown(self, skill_name: str, metric_type: MetricType) -> bool:
+        """Check if a skill/metric combination is in cooldown period."""
+        key = f"{skill_name}:{metric_type.value}"
+        last_trigger = self._cooldowns.get(key)
+        if not last_trigger:
+            return False
+
+        # Find the threshold for this metric type
+        for threshold in self.thresholds:
+            if threshold.metric_type == metric_type:
+                elapsed = (datetime.now() - last_trigger).total_seconds()
+                return elapsed < threshold.cooldown_seconds
+
+        return False
+
+
+class AdaptationManager:
+    """Manager for coordinating skill adaptation across multiple agents.
+
+    Provides centralized tracking of adaptation events and skill reloads.
+    """
+
+    def __init__(self, storage_dir: Path):
+        """Initialize adaptation manager.
+
+        Args:
+            storage_dir: Root directory for storing adaptation data
+        """
+        self.storage_dir = Path(storage_dir)
+        self._hooks: Dict[str, SkillAdaptationHook] = {}
+
+    def get_hook(
+        self,
+        run_id: str,
+        agent_id: str,
+        thresholds: Optional[List[AdaptationThreshold]] = None,
+    ) -> SkillAdaptationHook:
+        """Get or create an adaptation hook for an agent.
+
+        Args:
+            run_id: Run identifier
+            agent_id: Agent identifier
+            thresholds: Optional custom thresholds
+
+        Returns:
+            SkillAdaptationHook instance
+        """
+        key = f"{run_id}:{agent_id}"
+        if key not in self._hooks:
+            self._hooks[key] = SkillAdaptationHook(
+                storage_dir=self.storage_dir,
+                run_id=run_id,
+                agent_id=agent_id,
+                thresholds=thresholds,
+            )
+        return self._hooks[key]
+
+    def get_all_pending_warnings(self) -> List[AdaptationEvent]:
+        """Get all pending warnings from all hooks."""
+        warnings = []
+        for hook in self._hooks.values():
+            warnings.extend(hook.get_pending_warnings())
+        return warnings
+
+    def get_run_adaptations(self, run_id: str) -> List[AdaptationEvent]:
+        """Get all adaptation events for a run."""
+        events = []
+        for hook in self._hooks.values():
+            if hook.run_id == run_id:
+                events.extend(hook.get_recent_events())
+        return events
+
+
+# Global manager instance
+_adaptation_manager: Optional[AdaptationManager] = None
+
+
+def get_adaptation_manager(storage_dir: Optional[Path] = None) -> AdaptationManager:
+    """Get global adaptation manager instance.
+
+    Args:
+        storage_dir: Optional storage directory (required on first call)
+
+    Returns:
+        AdaptationManager instance
+    """
+    global _adaptation_manager
+    if _adaptation_manager is None:
+        if storage_dir is None:
+            raise ValueError("storage_dir required on first initialization")
+        _adaptation_manager = AdaptationManager(storage_dir)
+    return _adaptation_manager
+
+
+__all__ = [
+    "AdaptationAction",
+    "AdaptationThreshold",
+    "AdaptationEvent",
+    "SkillAdaptationHook",
+    "AdaptationManager",
+    "get_adaptation_manager",
+]
--- a/backend/agents/base/tool_guard.py
+++ b/backend/agents/base/tool_guard.py
@@ -0,0 +1,684 @@
+# -*- coding: utf-8 -*-
+"""ToolGuardMixin - Security interception for dangerous tool calls.
+
+Provides ``_acting`` and ``_reasoning`` overrides that intercept
+sensitive tool calls before execution, implementing the deny /
+guard / approve flow.
+
+Based on CoPaw's tool_guard_mixin.py design.
+"""
+from __future__ import annotations
+
+import asyncio
+import json
+import logging
+from dataclasses import dataclass, field
+from datetime import datetime
+from enum import Enum
+
+from typing import Any, Callable, Dict, Iterable, List, Optional, Set
+
+from agentscope.message import Msg
+from backend.runtime.manager import get_global_runtime_manager
+logger = logging.getLogger(__name__)
+
+class SeverityLevel(str, Enum):
+    """Risk severity level."""
+
+    LOW = "low"
+    MEDIUM = "medium"
+    HIGH = "high"
+    CRITICAL = "critical"
+
+
+class ApprovalStatus(str, Enum):
+    """Approval lifecycle state."""
+
+    PENDING = "pending"
+    APPROVED = "approved"
+    DENIED = "denied"
+    EXPIRED = "expired"
+
+
+class ToolFindingRecord:
+    """Internal representation of a guard finding."""
+
+    def __init__(self, severity: SeverityLevel, message: str, field: Optional[str] = None) -> None:
+        self.severity = severity
+        self.message = message
+        self.field = field
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "severity": self.severity.value,
+            "message": self.message,
+            "field": self.field,
+        }
+
+
+class ApprovalRecord:
+    """Stores the state of an approval request."""
+
+    def __init__(
+        self,
+        approval_id: str,
+        tool_name: str,
+        tool_input: Dict[str, Any],
+        agent_id: str,
+        workspace_id: str,
+        session_id: Optional[str] = None,
+        findings: Optional[List[ToolFindingRecord]] = None,
+    ) -> None:
+        self.approval_id = approval_id
+        self.tool_name = tool_name
+        self.tool_input = tool_input
+        self.agent_id = agent_id
+        self.workspace_id = workspace_id
+        self.session_id = session_id
+        self.status = ApprovalStatus.PENDING
+        self.findings = findings or []
+        self.created_at = datetime.utcnow()
+        self.resolved_at: Optional[datetime] = None
+        self.resolved_by: Optional[str] = None
+        self.metadata: Dict[str, Any] = {}
+        self.pending_request: "ToolApprovalRequest" | None = None
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "approval_id": self.approval_id,
+            "status": self.status.value,
+            "tool_name": self.tool_name,
+            "tool_input": self.tool_input,
+            "agent_id": self.agent_id,
+            "workspace_id": self.workspace_id,
+            "session_id": self.session_id,
+            "findings": [f.to_dict() for f in self.findings],
+            "created_at": self.created_at.isoformat(),
+            "resolved_at": self.resolved_at.isoformat() if self.resolved_at else None,
+            "resolved_by": self.resolved_by,
+        }
+
+
+class ToolGuardStore:
+    """Simple in-memory approval store for development/testing."""
+
+    def __init__(self) -> None:
+        self._records: Dict[str, ApprovalRecord] = {}
+        self._counter = 0
+
+    def next_id(self) -> str:
+        self._counter += 1
+        return f"approval_{self._counter:06d}"
+
+    def list(
+        self,
+        status: ApprovalStatus | None = None,
+        workspace_id: Optional[str] = None,
+        agent_id: Optional[str] = None,
+    ) -> Iterable[ApprovalRecord]:
+        for record in self._records.values():
+            if status and record.status != status:
+                continue
+            if workspace_id and record.workspace_id != workspace_id:
+                continue
+            if agent_id and record.agent_id != agent_id:
+                continue
+            yield record
+
+    def get(self, approval_id: str) -> Optional[ApprovalRecord]:
+        return self._records.get(approval_id)
+
+    def create_pending(
+        self,
+        tool_name: str,
+        tool_input: Dict[str, Any],
+        agent_id: str,
+        workspace_id: str,
+        session_id: Optional[str] = None,
+        findings: Optional[List[ToolFindingRecord]] = None,
+    ) -> ApprovalRecord:
+        record = ApprovalRecord(
+            approval_id=self.next_id(),
+            tool_name=tool_name,
+            tool_input=tool_input,
+            agent_id=agent_id,
+            workspace_id=workspace_id,
+            session_id=session_id,
+            findings=findings,
+        )
+        self._records[record.approval_id] = record
+        return record
+
+    def set_status(
+        self,
+        approval_id: str,
+        status: ApprovalStatus,
+        resolved_by: Optional[str] = None,
+        notify_request: bool = True,
+    ) -> ApprovalRecord:
+        record = self._records[approval_id]
+        if record.status == status:
+            return record
+
+        record.status = status
+        record.resolved_at = datetime.utcnow()
+        record.resolved_by = resolved_by
+        if notify_request and record.pending_request:
+            if status == ApprovalStatus.APPROVED:
+                record.pending_request.approve()
+            elif status == ApprovalStatus.DENIED:
+                record.pending_request.deny()
+        return record
+
+    def cancel(self, approval_id: str) -> None:
+        self._records.pop(approval_id, None)
+
+
+TOOL_GUARD_STORE = ToolGuardStore()
+
+
+def get_tool_guard_store() -> ToolGuardStore:
+    return TOOL_GUARD_STORE
+
+
+# Default tools that require approval
+DEFAULT_GUARDED_TOOLS: Set[str] = {
+    "execute_shell_command",
+    "write_file",
+    "edit_file",
+    "place_order",
+    "modify_position",
+    "delete_file",
+}
+
+# Default denied tools (cannot be approved)
+DEFAULT_DENIED_TOOLS: Set[str] = {
+    "execute_shell_command",  # Shell execution is dangerous
+}
+
+# Mark for tool guard denied messages
+TOOL_GUARD_DENIED_MARK = "tool_guard_denied"
+
+
+def default_findings_for_tool(tool_name: str) -> List[ToolFindingRecord]:
+    findings: List[ToolFindingRecord] = []
+    if tool_name in {"execute_trade", "modify_portfolio"}:
+        findings.append(
+            ToolFindingRecord(
+                severity=SeverityLevel.HIGH,
+                message=f"Tool '{tool_name}' touches portfolio state",
+            )
+        )
+    return findings
+
+
+class ToolApprovalRequest:
+    """Represents a pending tool approval request."""
+
+    def __init__(
+        self,
+        approval_id: str,
+        tool_name: str,
+        tool_input: Dict[str, Any],
+        tool_call_id: str,
+        session_id: Optional[str] = None,
+    ):
+        self.approval_id = approval_id
+        self.tool_name = tool_name
+        self.tool_input = tool_input
+        self.tool_call_id = tool_call_id
+        self.session_id = session_id
+        self.approved: Optional[bool] = None
+        self._event = asyncio.Event()
+
+    async def wait_for_approval(self, timeout: Optional[float] = None) -> bool:
+        """Wait for approval decision.
+
+        Args:
+            timeout: Maximum time to wait in seconds
+
+        Returns:
+            True if approved, False otherwise
+        """
+        try:
+            await asyncio.wait_for(self._event.wait(), timeout=timeout)
+        except asyncio.TimeoutError:
+            return False
+        return self.approved is True
+
+    def approve(self) -> None:
+        """Approve this request."""
+        self.approved = True
+        self._event.set()
+
+    def deny(self) -> None:
+        """Deny this request."""
+        self.approved = False
+        self._event.set()
+
+
+class ToolGuardMixin:
+    """Mixin that adds tool-guard interception to a ReActAgent.
+
+    At runtime this class is combined with ReActAgent via MRO,
+    so ``super()._acting`` and ``super()._reasoning`` resolve to
+    the concrete agent methods.
+
+    Usage:
+        class MyAgent(ToolGuardMixin, ReActAgent):
+            def __init__(self, ...):
+                super().__init__(...)
+                self._init_tool_guard()
+    """
+
+    def _init_tool_guard(
+        self,
+        guarded_tools: Optional[Set[str]] = None,
+        denied_tools: Optional[Set[str]] = None,
+        approval_timeout: float = 300.0,
+    ) -> None:
+        """Initialize tool guard.
+
+        Args:
+            guarded_tools: Set of tool names requiring approval
+            denied_tools: Set of tool names that are always denied
+            approval_timeout: Timeout for approval requests in seconds
+        """
+        self._guarded_tools = guarded_tools or DEFAULT_GUARDED_TOOLS.copy()
+        self._denied_tools = denied_tools or DEFAULT_DENIED_TOOLS.copy()
+        self._approval_timeout = approval_timeout
+        self._pending_approval: Optional[ToolApprovalRequest] = None
+        self._approval_callback: Optional[Callable[[ToolApprovalRequest], None]] = None
+        self._approval_lock = asyncio.Lock()
+
+    def set_approval_callback(
+        self,
+        callback: Callable[[ToolApprovalRequest], None],
+    ) -> None:
+        """Set callback for approval requests.
+
+        Args:
+            callback: Function called when approval is needed
+        """
+        self._approval_callback = callback
+
+    def _is_tool_guarded(self, tool_name: str) -> bool:
+        """Check if a tool requires approval.
+
+        Args:
+            tool_name: Name of the tool
+
+        Returns:
+            True if tool requires approval
+        """
+        return tool_name in self._guarded_tools
+
+    def _is_tool_denied(self, tool_name: str) -> bool:
+        """Check if a tool is always denied.
+
+        Args:
+            tool_name: Name of the tool
+
+        Returns:
+            True if tool is denied
+        """
+        return tool_name in self._denied_tools
+
+    def _last_tool_response_is_denied(self) -> bool:
+        """Check if the last message is a guard-denied tool result."""
+        if not hasattr(self, "memory") or not self.memory.content:
+            return False
+
+        msg, marks = self.memory.content[-1]
+        return TOOL_GUARD_DENIED_MARK in marks and msg.role == "system"
+
+    async def _cleanup_tool_guard_denied_messages(
+        self,
+        include_denial_response: bool = True,
+    ) -> None:
+        """Remove tool-guard denied messages from memory.
+
+        Args:
+            include_denial_response: Also remove the assistant's denial explanation
+        """
+        if not hasattr(self, "memory"):
+            return
+
+        ids_to_delete: list[str] = []
+        last_marked_idx = -1
+
+        for i, (msg, marks) in enumerate(self.memory.content):
+            if TOOL_GUARD_DENIED_MARK in marks:
+                ids_to_delete.append(msg.id)
+                last_marked_idx = i
+
+        if (
+            include_denial_response
+            and last_marked_idx >= 0
+            and last_marked_idx + 1 < len(self.memory.content)
+        ):
+            next_msg, _ = self.memory.content[last_marked_idx + 1]
+            if next_msg.role == "assistant":
+                ids_to_delete.append(next_msg.id)
+
+        if ids_to_delete:
+            removed = await self.memory.delete(ids_to_delete)
+            logger.info("Tool guard: cleaned up %d denied message(s)", removed)
+
+    async def _request_guard_approval(
+        self,
+        tool_name: str,
+        tool_input: Dict[str, Any],
+        tool_call_id: str,
+    ) -> bool:
+        """Request approval for a guarded tool call.
+
+        This method creates a ToolApprovalRequest and waits for
+        external approval via approve_guard_call() or deny_guard_call().
+
+        Args:
+            tool_name: Name of the tool
+            tool_input: Tool input parameters
+            tool_call_id: ID of the tool call
+
+        Returns:
+            True if approved, False otherwise
+        """
+        async with self._approval_lock:
+            record = TOOL_GUARD_STORE.create_pending(
+                tool_name=tool_name,
+                tool_input=tool_input,
+                agent_id=getattr(self, "agent_id", "unknown"),
+                workspace_id=getattr(self, "workspace_id", "default"),
+                session_id=getattr(self, "session_id", None),
+                findings=default_findings_for_tool(tool_name),
+            )
+
+            manager = get_global_runtime_manager()
+            if manager:
+                manager.register_pending_approval(
+                    record.approval_id,
+                    {
+                        "tool_name": record.tool_name,
+                        "agent_id": record.agent_id,
+                        "workspace_id": record.workspace_id,
+                        "session_id": record.session_id,
+                        "tool_input": record.tool_input,
+                    },
+                )
+
+            self._pending_approval = ToolApprovalRequest(
+                approval_id=record.approval_id,
+                tool_name=tool_name,
+                tool_input=tool_input,
+                tool_call_id=tool_call_id,
+                session_id=getattr(self, "session_id", None),
+            )
+            record.pending_request = self._pending_approval
+
+            # Notify via callback if set
+            if self._approval_callback:
+                self._approval_callback(self._pending_approval)
+
+            # Wait for approval (lock is released during wait, re-acquired after)
+            approval_request = self._pending_approval
+
+        # Wait for approval outside the lock to allow concurrent approval
+        approved = await approval_request.wait_for_approval(
+            timeout=self._approval_timeout
+        )
+
+        async with self._approval_lock:
+            if approval_request:
+                status = (
+                    ApprovalStatus.APPROVED
+                    if approval_request.approved is True
+                    else ApprovalStatus.DENIED
+                    if approval_request.approved is False
+                    else ApprovalStatus.EXPIRED
+                )
+                TOOL_GUARD_STORE.set_status(
+                    approval_request.approval_id,
+                    status,
+                    resolved_by="agent",
+                    notify_request=False,
+                )
+                manager = get_global_runtime_manager()
+                if manager:
+                    manager.resolve_pending_approval(
+                        approval_request.approval_id,
+                        resolved_by="agent",
+                        status=status.value,
+                    )
+
+            # Only clear if this is still the same request
+            if self._pending_approval is approval_request:
+                self._pending_approval = None
+
+        return approved
+
+    async def approve_guard_call(self, request_id: Optional[str] = None) -> bool:
+        """Approve a pending guard request.
+
+        This method is called externally to approve a tool call
+        that is waiting for approval.
+
+        Args:
+            request_id: Optional request ID to verify (not yet implemented)
+
+        Returns:
+            True if a request was approved, False if no pending request
+        """
+        async with self._approval_lock:
+            if self._pending_approval is None:
+                logger.warning("No pending approval request to approve")
+                return False
+
+            TOOL_GUARD_STORE.set_status(
+                self._pending_approval.approval_id,
+                ApprovalStatus.APPROVED,
+                resolved_by="agent",
+                notify_request=False,
+            )
+            manager = get_global_runtime_manager()
+            if manager:
+                manager.resolve_pending_approval(
+                    self._pending_approval.approval_id,
+                    resolved_by="agent",
+                    status=ApprovalStatus.APPROVED.value,
+                )
+            self._pending_approval.approve()
+            logger.info("Approved tool call: %s", self._pending_approval.tool_name)
+            return True
+
+    async def deny_guard_call(self, request_id: Optional[str] = None) -> bool:
+        """Deny a pending guard request.
+
+        This method is called externally to deny a tool call
+        that is waiting for approval.
+
+        Args:
+            request_id: Optional request ID to verify (not yet implemented)
+
+        Returns:
+            True if a request was denied, False if no pending request
+        """
+        async with self._approval_lock:
+            if self._pending_approval is None:
+                logger.warning("No pending approval request to deny")
+                return False
+
+            TOOL_GUARD_STORE.set_status(
+                self._pending_approval.approval_id,
+                ApprovalStatus.DENIED,
+                resolved_by="agent",
+                notify_request=False,
+            )
+            manager = get_global_runtime_manager()
+            if manager:
+                manager.resolve_pending_approval(
+                    self._pending_approval.approval_id,
+                    resolved_by="agent",
+                    status=ApprovalStatus.DENIED.value,
+                )
+            self._pending_approval.deny()
+            logger.info("Denied tool call: %s", self._pending_approval.tool_name)
+            return True
+
+    async def _acting(self, tool_call) -> dict | None:
+        """Intercept sensitive tool calls before execution.
+
+        1. If tool is in denied_tools, auto-deny unconditionally.
+        2. Check for a one-shot pre-approval.
+        3. If tool is in the guarded scope, request approval.
+        4. Otherwise, delegate to parent _acting.
+
+        Args:
+            tool_call: Tool call from the model
+
+        Returns:
+            Tool result dict or None
+        """
+        tool_name: str = tool_call.get("name", "")
+        tool_input: dict = tool_call.get("input", {})
+        tool_call_id: str = tool_call.get("id", "")
+
+        # Check if tool is denied
+        if tool_name and self._is_tool_denied(tool_name):
+            logger.warning("Tool '%s' is in the denied set, auto-denying", tool_name)
+            return await self._acting_auto_denied(tool_call, tool_name)
+
+        # Check if tool is guarded
+        if tool_name and self._is_tool_guarded(tool_name):
+            approved = await self._request_guard_approval(
+                tool_name=tool_name,
+                tool_input=tool_input,
+                tool_call_id=tool_call_id,
+            )
+
+            if not approved:
+                return await self._acting_with_denial(tool_call, tool_name)
+
+        # Call parent _acting
+        return await super()._acting(tool_call)  # type: ignore[misc]
+
+    async def _acting_auto_denied(
+        self,
+        tool_call: Dict[str, Any],
+        tool_name: str,
+    ) -> dict | None:
+        """Auto-deny a tool call without offering approval.
+
+        Args:
+            tool_call: Tool call from the model
+            tool_name: Name of the denied tool
+
+        Returns:
+            Denial result
+        """
+        from agentscope.message import ToolResultBlock
+
+        denied_text = (
+            f"⛔ **Tool Blocked / 工具已拦截**\n\n"
+            f"- Tool / 工具: `{tool_name}`\n"
+            f"- Reason / 原因: This tool is blocked for security reasons\n\n"
+            f"This tool is blocked and cannot be approved.\n"
+            f"该工具已被禁止，无法批准执行。"
+        )
+
+        tool_res_msg = Msg(
+            "system",
+            [
+                ToolResultBlock(
+                    type="tool_result",
+                    id=tool_call.get("id", ""),
+                    name=tool_name,
+                    output=[{"type": "text", "text": denied_text}],
+                ),
+            ],
+            "system",
+        )
+
+        await self.print(tool_res_msg, True)
+        await self.memory.add(tool_res_msg)
+        return None
+
+    async def _acting_with_denial(
+        self,
+        tool_call: Dict[str, Any],
+        tool_name: str,
+    ) -> dict | None:
+        """Deny the tool call after approval was rejected.
+
+        Args:
+            tool_call: Tool call from the model
+            tool_name: Name of the tool
+
+        Returns:
+            Denial result
+        """
+        from agentscope.message import ToolResultBlock
+
+        params_text = json.dumps(
+            tool_call.get("input", {}),
+            ensure_ascii=False,
+            indent=2,
+        )
+
+        denied_text = (
+            f"⚠️ **Tool Call Denied / 工具调用被拒绝**\n\n"
+            f"- Tool / 工具: `{tool_name}`\n"
+            f"- Parameters / 参数:\n"
+            f"```json\n{params_text}\n```\n\n"
+            f"The tool call was denied by the user or timed out.\n"
+            f"工具调用被用户拒绝或已超时。"
+        )
+
+        tool_res_msg = Msg(
+            "system",
+            [
+                ToolResultBlock(
+                    type="tool_result",
+                    id=tool_call.get("id", ""),
+                    name=tool_name,
+                    output=[{"type": "text", "text": denied_text}],
+                ),
+            ],
+            "system",
+        )
+
+        await self.print(tool_res_msg, True)
+        await self.memory.add(tool_res_msg, marks=TOOL_GUARD_DENIED_MARK)
+        return None
+
+    async def _reasoning(self, **kwargs) -> Msg:
+        """Short-circuit reasoning when awaiting guard approval.
+
+        If the last message was a guard denial, return a waiting message
+        instead of continuing reasoning.
+
+        Returns:
+            Response message
+        """
+        if self._last_tool_response_is_denied():
+            msg = Msg(
+                self.name,
+                "⏳ Waiting for approval / 等待审批...\n\n"
+                "Type `/approve` to approve, or send any message to deny.\n"
+                "输入 `/approve` 批准执行，或发送任意消息拒绝。",
+                "assistant",
+            )
+            await self.print(msg, True)
+            await self.memory.add(msg)
+            return msg
+
+        return await super()._reasoning(**kwargs)  # type: ignore[misc]
+
+
+__all__ = [
+    "ToolGuardMixin",
+    "ToolApprovalRequest",
+    "DEFAULT_GUARDED_TOOLS",
+    "DEFAULT_DENIED_TOOLS",
+    "TOOL_GUARD_DENIED_MARK",
+]
--- a/backend/agents/compat.py
+++ b/backend/agents/compat.py
@@ -0,0 +1,146 @@
+# -*- coding: utf-8 -*-
+"""
+Compatibility Layer - Adapters for legacy to EvoAgent migration.
+
+Provides:
+- LegacyAgentAdapter: Wraps old AnalystAgent to work with new interfaces
+- Migration utilities for gradual adoption
+"""
+from typing import Any, Dict, Optional
+
+from agentscope.message import Msg
+
+from .agent_core import EvoAgent
+
+
+class LegacyAgentAdapter:
+    """
+    Adapter to make legacy AnalystAgent compatible with EvoAgent interfaces.
+
+    This allows gradual migration by wrapping existing agents.
+    """
+
+    def __init__(self, legacy_agent: Any):
+        """
+        Initialize adapter.
+
+        Args:
+            legacy_agent: Legacy AnalystAgent instance
+        """
+        self._agent = legacy_agent
+        self.agent_id = getattr(legacy_agent, 'agent_id', getattr(legacy_agent, 'name', 'unknown'))
+        self.analyst_type = getattr(legacy_agent, 'analyst_type_key', None)
+
+    @property
+    def name(self) -> str:
+        """Get agent name."""
+        return getattr(self._agent, 'name', self.agent_id)
+
+    @property
+    def toolkit(self) -> Any:
+        """Get agent toolkit."""
+        return getattr(self._agent, 'toolkit', None)
+
+    @property
+    def model(self) -> Any:
+        """Get agent model."""
+        return getattr(self._agent, 'model', None)
+
+    @property
+    def memory(self) -> Any:
+        """Get agent memory."""
+        return getattr(self._agent, 'memory', None)
+
+    async def reply(self, x: Msg = None) -> Msg:
+        """
+        Delegate to legacy agent's reply method.
+
+        Args:
+            x: Input message
+
+        Returns:
+            Response message
+        """
+        return await self._agent.reply(x)
+
+    def reload_runtime_assets(self, active_skill_dirs: Optional[list] = None) -> None:
+        """
+        Reload runtime assets if supported.
+
+        Args:
+            active_skill_dirs: Optional list of active skill directories
+        """
+        if hasattr(self._agent, 'reload_runtime_assets'):
+            self._agent.reload_runtime_assets(active_skill_dirs)
+
+    def to_evo_agent(
+        self,
+        workspace_manager: Optional[Any] = None,
+        enable_tool_guard: bool = False,
+    ) -> EvoAgent:
+        """
+        Convert legacy agent to EvoAgent.
+
+        Args:
+            workspace_manager: Optional workspace manager
+            enable_tool_guard: Whether to enable tool guard
+
+        Returns:
+            New EvoAgent instance with same configuration
+        """
+        return EvoAgent(
+            agent_id=self.agent_id,
+            model=self.model,
+            formatter=getattr(self._agent, 'formatter', None),
+            toolkit=self.toolkit,
+            workspace_manager=workspace_manager,
+            config=getattr(self._agent, 'config', {}),
+            long_term_memory=getattr(self._agent, 'long_term_memory', None),
+            enable_tool_guard=enable_tool_guard,
+            sys_prompt=getattr(self._agent, '_sys_prompt', None),
+        )
+
+    def __getattr__(self, name: str) -> Any:
+        """Delegate unknown attributes to wrapped agent."""
+        return getattr(self._agent, name)
+
+
+def is_legacy_agent(agent: Any) -> bool:
+    """
+    Check if an agent is a legacy agent.
+
+    Args:
+        agent: Agent instance to check
+
+    Returns:
+        True if legacy agent
+    """
+    return hasattr(agent, 'analyst_type_key') and not isinstance(agent, EvoAgent)
+
+
+def adapt_agent(agent: Any) -> Any:
+    """
+    Wrap agent in adapter if it's a legacy agent.
+
+    Args:
+        agent: Agent instance
+
+    Returns:
+        Adapted agent or original if already EvoAgent
+    """
+    if is_legacy_agent(agent):
+        return LegacyAgentAdapter(agent)
+    return agent
+
+
+def adapt_agents(agents: list) -> list:
+    """
+    Wrap multiple agents in adapters.
+
+    Args:
+        agents: List of agent instances
+
+    Returns:
+        List of adapted agents
+    """
+    return [adapt_agent(agent) for agent in agents]
--- a/backend/agents/factory.py
+++ b/backend/agents/factory.py
@@ -0,0 +1,497 @@
+# -*- coding: utf-8 -*-
+"""Agent Factory - Dynamic creation and management of AgentConfigs."""
+
+import logging
+import shutil
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+import yaml
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class ModelConfig:
+    """Model configuration for an agent."""
+
+    model_name: str = "gpt-4o"
+    temperature: float = 0.7
+    max_tokens: int = 4096
+
+
+@dataclass
+class RoleConfig:
+    """Role configuration for an agent."""
+
+    name: str
+    description: str = ""
+    focus_areas: List[str] = None
+    constraints: List[str] = None
+
+    def __post_init__(self):
+        if self.focus_areas is None:
+            self.focus_areas = []
+        if self.constraints is None:
+            self.constraints = []
+
+
+class AgentConfig:
+    """Represents a configured agent instance (data class)."""
+
+    def __init__(
+        self,
+        agent_id: str,
+        agent_type: str,
+        workspace_id: str,
+        config_path: Path,
+        model_config: Optional[ModelConfig] = None,
+        role_config: Optional[RoleConfig] = None,
+    ):
+        self.agent_id = agent_id
+        self.agent_type = agent_type
+        self.workspace_id = workspace_id
+        self.config_path = config_path
+        self.model_config = model_config or ModelConfig()
+        self.role_config = role_config
+        self.agent_dir = config_path.parent
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Serialize agent to dictionary."""
+        return {
+            "agent_id": self.agent_id,
+            "agent_type": self.agent_type,
+            "workspace_id": self.workspace_id,
+            "config_path": str(self.config_path),
+            "agent_dir": str(self.agent_dir),
+            "model_config": {
+                "model_name": self.model_config.model_name,
+                "temperature": self.model_config.temperature,
+                "max_tokens": self.model_config.max_tokens,
+            },
+            "role_config": self.role_config.__dict__ if self.role_config else None,
+        }
+
+
+class AgentFactory:
+    """Factory for creating, cloning, and managing agents."""
+
+    # Default role templates by agent type
+    ROLE_TEMPLATES = {
+        "technical_analyst": {
+            "name": "Technical Analyst",
+            "description": "Analyze price patterns, trends, and technical indicators.",
+            "focus_areas": [
+                "Price action and chart patterns",
+                "Support and resistance levels",
+                "Technical indicators (RSI, MACD, Moving Averages)",
+                "Volume analysis",
+            ],
+            "constraints": [
+                "State clear signal, confidence, and invalidation conditions",
+                "Use available technical analysis tools",
+            ],
+        },
+        "fundamentals_analyst": {
+            "name": "Fundamentals Analyst",
+            "description": "Analyze company financials, earnings, and business metrics.",
+            "focus_areas": [
+                "Financial statements analysis",
+                "Earnings reports and guidance",
+                "Valuation metrics",
+                "Business model and competitive position",
+            ],
+            "constraints": [
+                "State clear signal, confidence, and invalidation conditions",
+                "Use available fundamental analysis tools",
+            ],
+        },
+        "sentiment_analyst": {
+            "name": "Sentiment Analyst",
+            "description": "Analyze market sentiment, news, and social signals.",
+            "focus_areas": [
+                "News sentiment analysis",
+                "Social media sentiment",
+                "Analyst ratings and price targets",
+                "Insider activity",
+            ],
+            "constraints": [
+                "State clear signal, confidence, and invalidation conditions",
+                "Use available sentiment analysis tools",
+            ],
+        },
+        "valuation_analyst": {
+            "name": "Valuation Analyst",
+            "description": "Perform valuation analysis and price target calculations.",
+            "focus_areas": [
+                "DCF and comparable valuation",
+                "Price target derivation",
+                "Margin of safety assessment",
+                "Risk-adjusted return expectations",
+            ],
+            "constraints": [
+                "State clear signal, confidence, and invalidation conditions",
+                "Use available valuation tools",
+            ],
+        },
+        "risk_manager": {
+            "name": "Risk Manager",
+            "description": "Quantify concentration, leverage, liquidity, and volatility risk.",
+            "focus_areas": [
+                "Portfolio concentration risk",
+                "Leverage and margin analysis",
+                "Liquidity assessment",
+                "Volatility and drawdown risk",
+            ],
+            "constraints": [
+                "Prioritize highest-severity risk first",
+                "State concrete limits and recommendations",
+                "Use available risk tools before issuing final memo",
+            ],
+        },
+        "portfolio_manager": {
+            "name": "Portfolio Manager",
+            "description": "Synthesize analyst and risk inputs into portfolio decisions.",
+            "focus_areas": [
+                "Position sizing and allocation",
+                "Risk-adjusted portfolio construction",
+                "Trade execution timing",
+                "Portfolio rebalancing",
+            ],
+            "constraints": [
+                "Be concise, capital-aware, and explicit about sizing rationale",
+                "Respect cash, margin, and concentration constraints",
+                "Consider all analyst inputs before decisions",
+            ],
+        },
+    }
+
+    def __init__(self, project_root: Optional[Path] = None):
+        """Initialize the agent factory.
+
+        Args:
+            project_root: Root directory of the project
+        """
+        self.project_root = project_root or Path(__file__).parent.parent.parent
+        self.workspaces_root = self.project_root / "workspaces"
+        self.template_dir = self.project_root / "backend" / "workspaces" / ".template"
+
+    def create_agent(
+        self,
+        agent_id: str,
+        agent_type: str,
+        workspace_id: str,
+        model_config: Optional[ModelConfig] = None,
+        role_config: Optional[RoleConfig] = None,
+        clone_from: Optional[str] = None,
+    ) -> AgentConfig:
+        """Create a new agent.
+
+        Args:
+            agent_id: Unique identifier for the agent
+            agent_type: Type of agent (e.g., "technical_analyst")
+            workspace_id: ID of the workspace to create agent in
+            model_config: Model configuration
+            role_config: Role configuration (auto-generated if None)
+            clone_from: Path to existing agent to clone from (optional)
+
+        Returns:
+            AgentConfig instance
+
+        Raises:
+            ValueError: If agent already exists or workspace doesn't exist
+        """
+        workspace_dir = self.workspaces_root / workspace_id
+        if not workspace_dir.exists():
+            raise ValueError(f"Workspace '{workspace_id}' does not exist")
+
+        agent_dir = workspace_dir / "agents" / agent_id
+        if agent_dir.exists():
+            raise ValueError(f"Agent '{agent_id}' already exists in workspace '{workspace_id}'")
+
+        # Create directory structure
+        agent_dir.mkdir(parents=True, exist_ok=True)
+        (agent_dir / "skills" / "active").mkdir(parents=True, exist_ok=True)
+        (agent_dir / "skills" / "local").mkdir(parents=True, exist_ok=True)
+        (agent_dir / "skills" / "installed").mkdir(parents=True, exist_ok=True)
+        (agent_dir / "skills" / "disabled").mkdir(parents=True, exist_ok=True)
+
+        # Copy template or clone existing agent
+        if clone_from:
+            self._clone_agent_files(clone_from, agent_dir, agent_id)
+        else:
+            self._copy_template(agent_dir, agent_id, agent_type)
+
+        # Generate role config if not provided
+        if role_config is None:
+            role_config = self._generate_role_config(agent_type)
+
+        # Generate ROLE.md
+        self._generate_role_md(agent_dir, role_config)
+
+        # Write agent.yaml
+        config_path = agent_dir / "agent.yaml"
+        self._write_agent_yaml(config_path, agent_id, agent_type, model_config)
+
+        return AgentConfig(
+            agent_id=agent_id,
+            agent_type=agent_type,
+            workspace_id=workspace_id,
+            config_path=config_path,
+            model_config=model_config,
+            role_config=role_config,
+        )
+
+    def delete_agent(self, agent_id: str, workspace_id: str) -> bool:
+        """Delete an agent and its workspace.
+
+        Args:
+            agent_id: ID of the agent to delete
+            workspace_id: ID of the workspace containing the agent
+
+        Returns:
+            True if deleted, False if agent didn't exist
+        """
+        agent_dir = self.workspaces_root / workspace_id / "agents" / agent_id
+        if not agent_dir.exists():
+            return False
+
+        shutil.rmtree(agent_dir)
+        return True
+
+    def clone_agent(
+        self,
+        source_agent_id: str,
+        source_workspace_id: str,
+        new_agent_id: str,
+        target_workspace_id: Optional[str] = None,
+        model_config: Optional[ModelConfig] = None,
+    ) -> AgentConfig:
+        """Clone an existing agent.
+
+        Args:
+            source_agent_id: ID of the agent to clone
+            source_workspace_id: Workspace containing the source agent
+            new_agent_id: ID for the new agent
+            target_workspace_id: Target workspace (defaults to source workspace)
+            model_config: Optional new model configuration
+
+        Returns:
+            AgentConfig instance for the cloned agent
+        """
+        target_workspace_id = target_workspace_id or source_workspace_id
+        source_dir = self.workspaces_root / source_workspace_id / "agents" / source_agent_id
+
+        if not source_dir.exists():
+            raise ValueError(f"Source agent '{source_agent_id}' not found")
+
+        # Load source agent config
+        source_config_path = source_dir / "agent.yaml"
+        source_config = {}
+        if source_config_path.exists():
+            with open(source_config_path, "r", encoding="utf-8") as f:
+                source_config = yaml.safe_load(f) or {}
+
+        agent_type = source_config.get("agent_type", "generic")
+
+        # Determine source path for cloning
+        clone_from = str(source_dir)
+
+        return self.create_agent(
+            agent_id=new_agent_id,
+            agent_type=agent_type,
+            workspace_id=target_workspace_id,
+            model_config=model_config,
+            clone_from=clone_from,
+        )
+
+    def list_agents(self, workspace_id: Optional[str] = None) -> List[Dict[str, Any]]:
+        """List all agents.
+
+        Args:
+            workspace_id: Optional workspace to filter by
+
+        Returns:
+            List of agent information dictionaries
+        """
+        agents = []
+
+        if workspace_id:
+            workspaces = [self.workspaces_root / workspace_id]
+        else:
+            if not self.workspaces_root.exists():
+                return agents
+            workspaces = [d for d in self.workspaces_root.iterdir() if d.is_dir()]
+
+        for workspace in workspaces:
+            agents_dir = workspace / "agents"
+            if not agents_dir.exists():
+                continue
+
+            for agent_dir in agents_dir.iterdir():
+                if not agent_dir.is_dir():
+                    continue
+
+                config_path = agent_dir / "agent.yaml"
+                if config_path.exists():
+                    try:
+                        with open(config_path, "r", encoding="utf-8") as f:
+                            config = yaml.safe_load(f) or {}
+
+                        agents.append({
+                            "agent_id": agent_dir.name,
+                            "workspace_id": workspace.name,
+                            "agent_type": config.get("agent_type", "unknown"),
+                            "config_path": str(config_path),
+                        })
+                    except Exception as e:
+                        logger.warning(f"Failed to load agent config {config_path}: {e}")
+
+        return agents
+
+    def _copy_template(
+        self,
+        agent_dir: Path,
+        agent_id: str,
+        agent_type: str,
+    ) -> None:
+        """Copy template files to agent directory.
+
+        Args:
+            agent_dir: Target agent directory
+            agent_id: ID of the agent
+            agent_type: Type of the agent
+        """
+        # Create default markdown files
+        default_files = {
+            "AGENTS.md": f"# Agent Guide\n\nDocument how {agent_id} should work, collaborate, and choose tools or skills.\n\n",
+            "SOUL.md": f"# Soul\n\nDescribe {agent_id}'s temperament, reasoning posture, and voice.\n\n",
+            "PROFILE.md": f"# Profile\n\nTrack {agent_id}'s long-lived investment style, preferences, and strengths.\n\n",
+            "MEMORY.md": f"# Memory\n\nStore durable lessons, heuristics, and reminders for {agent_id}.\n\n",
+            "HEARTBEAT.md": f"# Heartbeat\n\nOptional checklist for periodic review or self-reflection.\n\n",
+            "POLICY.md": f"# Policy\n\nOptional run-scoped constraints, limits, or strategy policy.\n\n",
+            "STYLE.md": f"# Style\n\nOptional run-scoped communication or reasoning style.\n\n",
+        }
+
+        for filename, content in default_files.items():
+            filepath = agent_dir / filename
+            if not filepath.exists():
+                filepath.write_text(content, encoding="utf-8")
+
+    def _clone_agent_files(self, source_path: str, target_dir: Path, new_agent_id: str) -> None:
+        """Clone files from an existing agent.
+
+        Args:
+            source_path: Path to source agent directory
+            target_dir: Target agent directory
+            new_agent_id: ID for the new agent
+        """
+        source_dir = Path(source_path)
+        if not source_dir.exists():
+            raise ValueError(f"Source path '{source_path}' does not exist")
+
+        # Copy markdown files
+        for md_file in source_dir.glob("*.md"):
+            target_file = target_dir / md_file.name
+            content = md_file.read_text(encoding="utf-8")
+            # Update agent references in content
+            source_name = source_dir.name
+            content = content.replace(source_name, new_agent_id)
+            target_file.write_text(content, encoding="utf-8")
+
+        # Copy skills directory structure (but not contents)
+        for skill_subdir in ["active", "local", "installed", "disabled"]:
+            source_skills = source_dir / "skills" / skill_subdir
+            if source_skills.exists():
+                target_skills = target_dir / "skills" / skill_subdir
+                target_skills.mkdir(parents=True, exist_ok=True)
+                # Copy skill files
+                for skill_file in source_skills.iterdir():
+                    if skill_file.is_file():
+                        shutil.copy2(skill_file, target_skills / skill_file.name)
+
+    def _generate_role_config(self, agent_type: str) -> RoleConfig:
+        """Generate role configuration for an agent type.
+
+        Args:
+            agent_type: Type of agent
+
+        Returns:
+            RoleConfig instance
+        """
+        template = self.ROLE_TEMPLATES.get(agent_type, {})
+        return RoleConfig(
+            name=template.get("name", agent_type.replace("_", " ").title()),
+            description=template.get("description", ""),
+            focus_areas=template.get("focus_areas", []),
+            constraints=template.get("constraints", []),
+        )
+
+    def _generate_role_md(self, agent_dir: Path, role_config: RoleConfig) -> None:
+        """Generate ROLE.md file.
+
+        Args:
+            agent_dir: Agent directory
+            role_config: Role configuration
+        """
+        lines = [f"# {role_config.name}", ""]
+
+        if role_config.description:
+            lines.extend([role_config.description, ""])
+
+        if role_config.focus_areas:
+            lines.extend(["## Focus Areas", ""])
+            for area in role_config.focus_areas:
+                lines.append(f"- {area}")
+            lines.append("")
+
+        if role_config.constraints:
+            lines.extend(["## Constraints", ""])
+            for constraint in role_config.constraints:
+                lines.append(f"- {constraint}")
+            lines.append("")
+
+        content = "\n".join(lines)
+        (agent_dir / "ROLE.md").write_text(content, encoding="utf-8")
+
+    def _write_agent_yaml(
+        self,
+        config_path: Path,
+        agent_id: str,
+        agent_type: str,
+        model_config: Optional[ModelConfig] = None,
+    ) -> None:
+        """Write agent.yaml configuration file.
+
+        Args:
+            config_path: Path to write configuration
+            agent_id: Agent ID
+            agent_type: Agent type
+            model_config: Optional model configuration
+        """
+        config = {
+            "agent_id": agent_id,
+            "agent_type": agent_type,
+            "prompt_files": [
+                "SOUL.md",
+                "PROFILE.md",
+                "AGENTS.md",
+                "POLICY.md",
+                "MEMORY.md",
+            ],
+            "enabled_skills": [],
+            "disabled_skills": [],
+            "active_tool_groups": [],
+            "disabled_tool_groups": [],
+        }
+
+        if model_config:
+            config["model"] = {
+                "name": model_config.model_name,
+                "temperature": model_config.temperature,
+                "max_tokens": model_config.max_tokens,
+            }
+
+        with open(config_path, "w", encoding="utf-8") as f:
+            yaml.safe_dump(config, f, allow_unicode=True, sort_keys=False)
--- a/backend/agents/portfolio_manager.py
+++ b/backend/agents/portfolio_manager.py
@@ -4,7 +4,8 @@ Portfolio Manager Agent - Based on AgentScope ReActAgent
 Responsible for decision-making (NOT trade execution)
 """

-from typing import Any, Dict, Optional
+from pathlib import Path
+from typing import Any, Dict, Optional, Callable

 from agentscope.agent import ReActAgent
 from agentscope.memory import InMemoryMemory, LongTermMemoryBase
@@ -12,9 +13,9 @@ from agentscope.message import Msg, TextBlock
 from agentscope.tool import Toolkit, ToolResponse

 from ..utils.progress import progress
-from .prompt_loader import PromptLoader
-
-_prompt_loader = PromptLoader()
+from .prompt_factory import build_agent_system_prompt, clear_prompt_factory_cache
+from .team_pipeline_config import update_active_analysts
+from ..config.constants import ANALYST_TYPES


 class PMAgent(ReActAgent):
@@ -36,24 +37,54 @@ class PMAgent(ReActAgent):
        margin_requirement: float = 0.25,
        config: Optional[Dict[str, Any]] = None,
        long_term_memory: Optional[LongTermMemoryBase] = None,
+        toolkit_factory: Any = None,
+        toolkit_factory_kwargs: Optional[Dict[str, Any]] = None,
+        toolkit: Optional[Toolkit] = None,
    ):
-        self.config = config or {}
+        object.__setattr__(self, "config", config or {})

        # Portfolio state
-        self.portfolio = {
-            "cash": initial_cash,
-            "positions": {},
-            "margin_used": 0.0,
-            "margin_requirement": margin_requirement,
-        }
+        object.__setattr__(
+            self,
+            "portfolio",
+            {
+                "cash": initial_cash,
+                "positions": {},
+                "margin_used": 0.0,
+                "margin_requirement": margin_requirement,
+            },
+        )

        # Decisions made in current cycle
-        self._decisions: Dict[str, Dict] = {}
+        object.__setattr__(self, "_decisions", {})
+        toolkit_factory_kwargs = toolkit_factory_kwargs or {}
+        object.__setattr__(self, "_toolkit_factory", toolkit_factory)
+        object.__setattr__(
+            self,
+            "_toolkit_factory_kwargs",
+            toolkit_factory_kwargs,
+        )
+        object.__setattr__(self, "_create_team_agent_cb", None)
+        object.__setattr__(self, "_remove_team_agent_cb", None)

-        # Create toolkit
-        toolkit = self._create_toolkit()
+        # Create toolkit after local state is ready so bound tool methods can be registered.
+        if toolkit is None:
+            if toolkit_factory is not None:
+                toolkit = toolkit_factory(
+                    name,
+                    self.config.get("config_name", "default"),
+                    owner=self,
+                    **toolkit_factory_kwargs,
+                )
+            else:
+                toolkit = self._create_toolkit()
+        object.__setattr__(self, "toolkit", toolkit)

-        sys_prompt = _prompt_loader.load_prompt("portfolio_manager", "system")
+        sys_prompt = build_agent_system_prompt(
+            agent_id=name,
+            config_name=self.config.get("config_name", "default"),
+            toolkit=self.toolkit,
+        )

        kwargs = {
            "name": name,
@@ -126,6 +157,107 @@ class PMAgent(ReActAgent):
            ],
        )

+    def _add_team_analyst(self, agent_id: str) -> ToolResponse:
+        """Add one analyst to active discussion team."""
+        config_name = self.config.get("config_name", "default")
+        project_root = Path(__file__).resolve().parents[2]
+        active = update_active_analysts(
+            project_root=project_root,
+            config_name=config_name,
+            available_analysts=list(ANALYST_TYPES.keys()),
+            add=[agent_id],
+        )
+        return ToolResponse(
+            content=[
+                TextBlock(
+                    type="text",
+                    text=(
+                        f"Active analyst team updated. Added: {agent_id}. "
+                        f"Current active analysts: {', '.join(active)}"
+                    ),
+                ),
+            ],
+        )
+
+    def _remove_team_analyst(self, agent_id: str) -> ToolResponse:
+        """Remove one analyst from active discussion team."""
+        callback_msg = ""
+        callback = self._remove_team_agent_cb
+        if callback is not None:
+            callback_msg = callback(agent_id=agent_id)
+
+        config_name = self.config.get("config_name", "default")
+        project_root = Path(__file__).resolve().parents[2]
+        active = update_active_analysts(
+            project_root=project_root,
+            config_name=config_name,
+            available_analysts=list(ANALYST_TYPES.keys()),
+            remove=[agent_id],
+        )
+        return ToolResponse(
+            content=[
+                TextBlock(
+                    type="text",
+                    text=(
+                        f"Active analyst team updated. Removed: {agent_id}. "
+                        f"Current active analysts: {', '.join(active)}"
+                        + (f" | {callback_msg}" if callback_msg else "")
+                    ),
+                ),
+            ],
+        )
+
+    def _set_active_analysts(self, agent_ids: str) -> ToolResponse:
+        """Set active analysts from comma-separated agent ids."""
+        requested = [
+            item.strip() for item in str(agent_ids or "").split(",") if item.strip()
+        ]
+        config_name = self.config.get("config_name", "default")
+        project_root = Path(__file__).resolve().parents[2]
+        active = update_active_analysts(
+            project_root=project_root,
+            config_name=config_name,
+            available_analysts=list(ANALYST_TYPES.keys()),
+            set_to=requested,
+        )
+        return ToolResponse(
+            content=[
+                TextBlock(
+                    type="text",
+                    text=f"Active analyst team set to: {', '.join(active)}",
+                ),
+            ],
+        )
+
+    def _create_team_analyst(self, agent_id: str, analyst_type: str) -> ToolResponse:
+        """Create a runtime analyst instance and activate it."""
+        callback = self._create_team_agent_cb
+        if callback is None:
+            return ToolResponse(
+                content=[
+                    TextBlock(
+                        type="text",
+                        text="Runtime agent creation is not available in current pipeline.",
+                    ),
+                ],
+            )
+        result = callback(agent_id=agent_id, analyst_type=analyst_type)
+        return ToolResponse(
+            content=[
+                TextBlock(type="text", text=result),
+            ],
+        )
+
+    def set_team_controller(
+        self,
+        *,
+        create_agent_callback: Optional[Callable[..., str]] = None,
+        remove_agent_callback: Optional[Callable[..., str]] = None,
+    ) -> None:
+        """Inject runtime team lifecycle callbacks from pipeline."""
+        object.__setattr__(self, "_create_team_agent_cb", create_agent_callback)
+        object.__setattr__(self, "_remove_team_agent_cb", remove_agent_callback)
+
    async def reply(self, x: Msg = None) -> Msg:
        """
        Make investment decisions
@@ -186,3 +318,71 @@ class PMAgent(ReActAgent):
    def update_portfolio(self, portfolio: Dict[str, Any]):
        """Update portfolio after external execution"""
        self.portfolio.update(portfolio)
+
+    def _has_open_positions(self) -> bool:
+        """Return whether the current portfolio still has non-zero positions."""
+        for position in self.portfolio.get("positions", {}).values():
+            if position.get("long", 0) or position.get("short", 0):
+                return True
+        return False
+
+    def can_apply_initial_cash(self) -> bool:
+        """Only allow cash rebasing before any positions or margin exist."""
+        return (
+            not self._has_open_positions()
+            and float(self.portfolio.get("margin_used", 0.0) or 0.0) == 0.0
+        )
+
+    def apply_runtime_portfolio_config(
+        self,
+        *,
+        margin_requirement: Optional[float] = None,
+        initial_cash: Optional[float] = None,
+    ) -> Dict[str, bool]:
+        """Apply safe run-time portfolio config updates."""
+        result = {
+            "margin_requirement": False,
+            "initial_cash": False,
+        }
+
+        if margin_requirement is not None:
+            self.portfolio["margin_requirement"] = float(margin_requirement)
+            result["margin_requirement"] = True
+
+        if initial_cash is not None and self.can_apply_initial_cash():
+            self.portfolio["cash"] = float(initial_cash)
+            result["initial_cash"] = True
+
+        return result
+
+    def reload_runtime_assets(self, active_skill_dirs: Optional[list] = None) -> None:
+        """Reload toolkit and system prompt from current run assets."""
+        from .toolkit_factory import create_agent_toolkit
+
+        clear_prompt_factory_cache()
+        toolkit_factory = self._toolkit_factory or create_agent_toolkit
+        toolkit_kwargs = dict(self._toolkit_factory_kwargs)
+        if active_skill_dirs is not None:
+            toolkit_kwargs["active_skill_dirs"] = active_skill_dirs
+
+        self.toolkit = toolkit_factory(
+            self.name,
+            self.config.get("config_name", "default"),
+            owner=self,
+            **toolkit_kwargs,
+        )
+        self._apply_runtime_sys_prompt(
+            build_agent_system_prompt(
+                agent_id=self.name,
+                config_name=self.config.get("config_name", "default"),
+                toolkit=self.toolkit,
+            ),
+        )
+
+    def _apply_runtime_sys_prompt(self, sys_prompt: str) -> None:
+        """Update the prompt used by future turns and the cached system msg."""
+        self._sys_prompt = sys_prompt
+        for msg, _marks in self.memory.content:
+            if getattr(msg, "role", None) == "system":
+                msg.content = sys_prompt
+                break
--- a/backend/agents/prompt_factory.py
+++ b/backend/agents/prompt_factory.py
@@ -0,0 +1,193 @@
+# -*- coding: utf-8 -*-
+"""Assemble system prompts from base prompts, run assets, and toolkit context."""
+
+from pathlib import Path
+from typing import Any, Optional
+
+from .agent_workspace import load_agent_workspace_config
+from backend.config.bootstrap_config import get_bootstrap_config_for_run
+from .prompt_loader import get_prompt_loader
+from .skills_manager import SkillsManager
+
+_prompt_loader = get_prompt_loader()
+
+
+def _read_file_if_exists(path: Path) -> str:
+    if not path.exists() or not path.is_file():
+        return ""
+    return path.read_text(encoding="utf-8").strip()
+
+
+def _append_section(parts: list[str], title: str, content: str) -> None:
+    content = content.strip()
+    if content:
+        parts.append(f"## {title}\n{content}")
+
+
+def _build_skill_metadata_summary(skills_manager: SkillsManager, config_name: str, agent_id: str) -> str:
+    """Create a compact summary of active skills for prompt routing."""
+    metadata_items = skills_manager.list_active_skill_metadata(config_name, agent_id)
+    if not metadata_items:
+        return ""
+
+    lines: list[str] = [
+        "You can use the following active skills. Prefer the most relevant one, then read its SKILL.md if needed for detailed workflow:",
+    ]
+    for item in metadata_items:
+        parts = [f"- `{item.skill_name}`"]
+        if item.description:
+            parts.append(item.description)
+        if item.version:
+            parts.append(f"version: {item.version}")
+        parts.append(f"path: {item.path}")
+        lines.append(" | ".join(parts))
+    return "\n".join(lines)
+
+
+def build_agent_system_prompt(
+    agent_id: str,
+    config_name: str,
+    toolkit: Any,
+    analyst_type: Optional[str] = None,
+) -> str:
+    """Build the final system prompt for an agent.
+
+    Always reads fresh from disk — no caching.
+    """
+    # Clear any cached templates before building (CoPaw-style, no caching)
+    _prompt_loader.clear_cache()
+
+    sections: list[str] = []
+    canonical_agent_id = (
+        "portfolio_manager"
+        if "portfolio" in agent_id
+        else "risk_manager"
+        if "risk" in agent_id and not analyst_type
+        else agent_id
+    )
+
+    if analyst_type:
+        personas_config = _prompt_loader.load_yaml_config(
+            "analyst",
+            "personas",
+        )
+        persona = personas_config.get(analyst_type, {})
+        focus_text = "\n".join(
+            f"- {item}" for item in persona.get("focus", [])
+        )
+        description = persona.get("description", "").strip()
+        base_prompt = _prompt_loader.load_prompt(
+            "analyst",
+            "system",
+            variables={
+                "analyst_type": persona.get("name", analyst_type),
+                "focus": focus_text,
+                "description": description,
+            },
+        )
+    elif agent_id == "portfolio_manager":
+        base_prompt = _prompt_loader.load_prompt(
+            "portfolio_manager",
+            "system",
+        )
+    elif canonical_agent_id == "portfolio_manager":
+        base_prompt = _prompt_loader.load_prompt(
+            "portfolio_manager",
+            "system",
+        )
+    elif agent_id == "risk_manager":
+        base_prompt = _prompt_loader.load_prompt(
+            "risk_manager",
+            "system",
+        )
+    elif canonical_agent_id == "risk_manager":
+        base_prompt = _prompt_loader.load_prompt(
+            "risk_manager",
+            "system",
+        )
+    else:
+        raise ValueError(f"Unsupported agent prompt build for: {agent_id}")
+
+    sections.append(base_prompt.strip())
+
+    skills_manager = SkillsManager()
+    asset_dir = skills_manager.get_agent_asset_dir(config_name, agent_id)
+    asset_dir.mkdir(parents=True, exist_ok=True)
+    agent_config = load_agent_workspace_config(asset_dir / "agent.yaml")
+    bootstrap_config = get_bootstrap_config_for_run(
+        skills_manager.project_root,
+        config_name,
+    )
+
+    _append_section(
+        sections,
+        "Bootstrap",
+        bootstrap_config.prompt_body,
+    )
+
+    prompt_files = agent_config.prompt_files or [
+        "SOUL.md",
+        "PROFILE.md",
+        "AGENTS.md",
+        "POLICY.md",
+        "MEMORY.md",
+    ]
+    included_files = set(prompt_files)
+    title_map = {
+        "SOUL.md": "Soul",
+        "PROFILE.md": "Profile",
+        "AGENTS.md": "Agent Guide",
+        "POLICY.md": "Policy",
+        "MEMORY.md": "Memory",
+        "HEARTBEAT.md": "Heartbeat",
+        "ROLE.md": "Role",
+        "STYLE.md": "Style",
+    }
+    for filename in prompt_files:
+        _append_section(
+            sections,
+            title_map.get(filename, filename),
+            _read_file_if_exists(asset_dir / filename),
+        )
+
+    if "ROLE.md" not in included_files:
+        _append_section(
+            sections,
+            "Role",
+            _read_file_if_exists(asset_dir / "ROLE.md"),
+        )
+    if "STYLE.md" not in included_files:
+        _append_section(
+            sections,
+            "Style",
+            _read_file_if_exists(asset_dir / "STYLE.md"),
+        )
+    if "POLICY.md" not in included_files:
+        _append_section(
+            sections,
+            "Policy",
+            _read_file_if_exists(asset_dir / "POLICY.md"),
+        )
+
+    skill_prompt = toolkit.get_agent_skill_prompt()
+    if skill_prompt:
+        _append_section(sections, "Skills", str(skill_prompt))
+
+    metadata_summary = _build_skill_metadata_summary(
+        skills_manager=skills_manager,
+        config_name=config_name,
+        agent_id=agent_id,
+    )
+    if metadata_summary:
+        _append_section(sections, "Active Skill Catalog", metadata_summary)
+
+    activated_notes = toolkit.get_activated_notes()
+    if activated_notes:
+        _append_section(sections, "Tool Usage Notes", str(activated_notes))
+
+    return "\n\n".join(section for section in sections if section.strip())
+
+
+def clear_prompt_factory_cache() -> None:
+    """Clear cached prompt and YAML templates before hot reload."""
+    _prompt_loader.clear_cache()
--- a/backend/agents/prompt_loader.py
+++ b/backend/agents/prompt_loader.py
@@ -10,6 +10,17 @@ from typing import Any, Dict, Optional

 import yaml

+# Singleton instance
+_prompt_loader_instance: Optional["PromptLoader"] = None
+
+
+def get_prompt_loader() -> "PromptLoader":
+    """Get the singleton PromptLoader instance."""
+    global _prompt_loader_instance
+    if _prompt_loader_instance is None:
+        _prompt_loader_instance = PromptLoader()
+    return _prompt_loader_instance
+

 class PromptLoader:
    """Unified Prompt loader"""
@@ -27,10 +38,6 @@ class PromptLoader:
        else:
            self.prompts_dir = Path(prompts_dir)

-        # Cache loaded prompts
-        self._prompt_cache: Dict[str, str] = {}
-        self._yaml_cache: Dict[str, Dict] = {}
-
    def load_prompt(
        self,
        agent_type: str,
@@ -38,37 +45,20 @@ class PromptLoader:
        variables: Optional[Dict[str, Any]] = None,
    ) -> str:
        """
-        Load and render Prompt
+        Load and render Prompt.

-        Args:
-            agent_type: Agent type (analyst, portfolio_manager, risk_manager)
-            prompt_name: Prompt file name (without extension)
-            variables: Variable dictionary for rendering Prompt
-
-        Returns:
-            Rendered prompt string
-
-        Examples:
-            loader = PromptLoader()
-            prompt = loader.load_prompt("analyst", "tool_selection",
-            {"analyst_persona": "Technical Analyst"})
+        No caching — always reads fresh from disk (CoPaw-style).
        """
-        cache_key = f"{agent_type}/{prompt_name}"
+        prompt_path = self.prompts_dir / agent_type / f"{prompt_name}.md"

-        # Try to load from cache
-        if cache_key not in self._prompt_cache:
-            prompt_path = self.prompts_dir / agent_type / f"{prompt_name}.md"
+        if not prompt_path.exists():
+            raise FileNotFoundError(
+                f"Prompt file not found: {prompt_path}\n"
+                f"Please create the prompt file or check the path.",
+            )

-            if not prompt_path.exists():
-                raise FileNotFoundError(
-                    f"Prompt file not found: {prompt_path}\n"
-                    f"Please create the prompt file or check the path.",
-                )
-
-            with open(prompt_path, "r", encoding="utf-8") as f:
-                self._prompt_cache[cache_key] = f.read()
-
-        prompt_template = self._prompt_cache[cache_key]
+        with open(prompt_path, "r", encoding="utf-8") as f:
+            prompt_template = f.read()

        # If variables provided, use simple string replacement
        if variables:
@@ -76,8 +66,6 @@ class PromptLoader:
        else:
            rendered = prompt_template

-        # Smart escaping: escape braces in JSON code blocks
-        # rendered = self._escape_json_braces(rendered)
        return rendered

    def _render_template(
@@ -140,45 +128,26 @@ class PromptLoader:
        config_name: str,
    ) -> Dict[str, Any]:
        """
-        Load YAML configuration file
+        Load YAML configuration file.

-        Args:
-            agent_type: Agent type
-            config_name: Configuration file name (without extension)
-
-        Returns:
-            Configuration dictionary
-
-        Examples:
-            >>> loader = PromptLoader()
-            >>> config = loader.load_yaml_config("analyst", "personas")
+        No caching — always reads fresh from disk (CoPaw-style).
        """
-        cache_key = f"{agent_type}/{config_name}"
+        yaml_path = self.prompts_dir / agent_type / f"{config_name}.yaml"

-        if cache_key not in self._yaml_cache:
-            yaml_path = self.prompts_dir / agent_type / f"{config_name}.yaml"
+        if not yaml_path.exists():
+            raise FileNotFoundError(f"YAML config not found: {yaml_path}")

-            if not yaml_path.exists():
-                raise FileNotFoundError(f"YAML config not found: {yaml_path}")
-
-            with open(yaml_path, "r", encoding="utf-8") as f:
-                self._yaml_cache[cache_key] = yaml.safe_load(f)
-
-        return self._yaml_cache[cache_key]
+        with open(yaml_path, "r", encoding="utf-8") as f:
+            return yaml.safe_load(f) or {}

    def clear_cache(self):
-        """Clear cache (for hot reload)"""
-        self._prompt_cache.clear()
-        self._yaml_cache.clear()
+        """No-op — caching removed (CoPaw-style, always fresh reads)."""
+        pass

    def reload_prompt(self, agent_type: str, prompt_name: str):
-        """Reload specified prompt (force cache refresh)"""
-        cache_key = f"{agent_type}/{prompt_name}"
-        if cache_key in self._prompt_cache:
-            del self._prompt_cache[cache_key]
+        """No-op — caching removed."""
+        pass

    def reload_config(self, agent_type: str, config_name: str):
-        """Reload specified configuration (force cache refresh)"""
-        cache_key = f"{agent_type}/{config_name}"
-        if cache_key in self._yaml_cache:
-            del self._yaml_cache[cache_key]
+        """No-op — caching removed."""
+        pass
--- a/backend/agents/prompts/init.py
+++ b/backend/agents/prompts/init.py
@@ -0,0 +1,19 @@
+# -*- coding: utf-8 -*-
+"""Prompt building utilities for EvoAgent.
+
+This module provides prompt construction from workspace markdown files
+with YAML frontmatter support.
+"""
+from .builder import (
+    PromptBuilder,
+    build_system_prompt_from_workspace,
+    build_bootstrap_guidance,
+    DEFAULT_SYS_PROMPT,
+)
+
+__all__ = [
+    "PromptBuilder",
+    "build_system_prompt_from_workspace",
+    "build_bootstrap_guidance",
+    "DEFAULT_SYS_PROMPT",
+]
--- a/backend/agents/prompts/analyst/personas.yaml
+++ b/backend/agents/prompts/analyst/personas.yaml
@@ -8,12 +8,6 @@ fundamentals_analyst:
    - "管理层质量和公司治理"
    - "行业地位和市场份额"
    - "长期投资价值评估"
-  tools:
-    - "analyze_profitability"
-    - "analyze_growth"
-    - "analyze_financial_health"
-    - "analyze_valuation_ratios"
-    - "analyze_efficiency_ratios"
  description: |
    作为基本面分析师，你专注于：
    - 公司财务健康状况和盈利能力
@@ -39,11 +33,6 @@ technical_analyst:
    - 支撑/阻力位和关键价格点
    - 中短期交易机会
    你倾向于选择能够捕捉价格动态和市场趋势的工具，更偏好技术分析类工具。
-  tools:
-    - "analyze_trend_following"
-    - "analyze_momentum"
-    - "analyze_mean_reversion"
-    - "analyze_volatility"

 sentiment_analyst:
  name: "情绪分析师"
@@ -61,10 +50,6 @@ sentiment_analyst:
    - 投资者恐慌和贪婪情绪
    - 市场预期和心理因素
    你倾向于选择能够反映市场情绪和投资者行为的工具，更偏好情绪和行为类工具。
-  tools:
-    - "analyze_news_sentiment"
-    - "analyze_insider_trading"
-
 valuation_analyst:
  name: "估值分析师"
  focus:
@@ -81,12 +66,6 @@ valuation_analyst:
    - 相对估值和绝对估值
    - 投资安全边际评估
    你倾向于选择能够准确计算公司价值的工具，更偏好估值模型和基本面工具。
-  tools:
-    - "dcf_valuation_analysis"
-    - "owner_earnings_valuation_analysis"
-    - "ev_ebitda_valuation_analysis"
-    - "residual_income_valuation_analysis"
-
 comprehensive_analyst:
  name: "综合分析师"
  focus:
@@ -103,15 +82,3 @@ comprehensive_analyst:
    - 提供全面的投资建议
    - 适应不同市场环境
    你会根据具体情况灵活选择各类工具，追求分析的全面性和准确性。
-  tools:
-    - "analyze_profitability"
-    - "analyze_growth"
-    - "analyze_financial_health"
-    - "analyze_valuation_ratios"
-    - "analyze_efficiency_ratios"
-    - "analyze_trend_following"
-    - "analyze_momentum"
-    - "analyze_mean_reversion"
-    - "analyze_volatility"
-    - "analyze_news_sentiment"
-    - "analyze_insider_trading"
--- a/backend/agents/prompts/builder.py
+++ b/backend/agents/prompts/builder.py
@@ -0,0 +1,305 @@
+# -*- coding: utf-8 -*-
+"""PromptBuilder for constructing system prompts from workspace markdown files.
+
+Based on CoPaw design - loads AGENTS.md, SOUL.md, PROFILE.md, etc. from
+agent workspace directories with YAML frontmatter support.
+"""
+from __future__ import annotations
+
+import logging
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+import yaml
+
+logger = logging.getLogger(__name__)
+
+DEFAULT_SYS_PROMPT = """You are a helpful trading analysis assistant."""
+
+
+class PromptBuilder:
+    """Builder for constructing system prompts from markdown files.
+
+    Loads markdown configuration files from agent workspace directories,
+    supporting YAML frontmatter for metadata extraction.
+    """
+
+    DEFAULT_FILES = [
+        "AGENTS.md",
+        "SOUL.md",
+        "PROFILE.md",
+        "ROLE.md",
+        "POLICY.md",
+        "MEMORY.md",
+        "HEARTBEAT.md",
+        "STYLE.md",
+    ]
+
+    TITLE_MAP: Dict[str, str] = {
+        "AGENTS.md": "Agent Guide",
+        "SOUL.md": "Soul",
+        "PROFILE.md": "Profile",
+        "ROLE.md": "Role",
+        "POLICY.md": "Policy",
+        "MEMORY.md": "Memory",
+        "HEARTBEAT.md": "Heartbeat",
+        "STYLE.md": "Style",
+        "BOOTSTRAP.md": "Bootstrap",
+    }
+
+    def __init__(
+        self,
+        workspace_dir: Path,
+        enabled_files: Optional[List[str]] = None,
+    ):
+        """Initialize prompt builder.
+
+        Args:
+            workspace_dir: Directory containing markdown configuration files
+            enabled_files: List of filenames to load (if None, uses defaults)
+        """
+        self.workspace_dir = Path(workspace_dir)
+        self.enabled_files = enabled_files or self.DEFAULT_FILES.copy()
+        self._prompt_parts: List[str] = []
+        self._metadata: Dict[str, Any] = {}
+        self.loaded_count = 0
+
+    def _load_file(self, filename: str) -> tuple[str, Optional[Dict[str, Any]]]:
+        """Load a single markdown file with YAML frontmatter support.
+
+        Args:
+            filename: Name of the file to load
+
+        Returns:
+            Tuple of (content, metadata dict or None)
+        """
+        file_path = self.workspace_dir / filename
+
+        if not file_path.exists():
+            logger.debug("File %s not found in %s, skipping", filename, self.workspace_dir)
+            return "", None
+
+        try:
+            raw_content = file_path.read_text(encoding="utf-8").strip()
+
+            if not raw_content:
+                logger.debug("Skipped empty file: %s", filename)
+                return "", None
+
+            content, metadata = self._parse_frontmatter(raw_content)
+
+            if content:
+                self.loaded_count += 1
+                logger.debug("Loaded %s (metadata: %s)", filename, bool(metadata))
+
+            return content, metadata
+
+        except Exception as e:
+            logger.warning("Failed to read file %s: %s, skipping", filename, e)
+            return "", None
+
+    def _parse_frontmatter(self, raw_content: str) -> tuple[str, Optional[Dict[str, Any]]]:
+        """Parse YAML frontmatter from markdown content.
+
+        Args:
+            raw_content: Raw file content
+
+        Returns:
+            Tuple of (content without frontmatter, metadata dict or None)
+        """
+        if not raw_content.startswith("---"):
+            return raw_content, None
+
+        parts = raw_content.split("---", 2)
+        if len(parts) < 3:
+            return raw_content, None
+
+        frontmatter = parts[1].strip()
+        content = parts[2].strip()
+
+        try:
+            metadata = yaml.safe_load(frontmatter) or {}
+            if not isinstance(metadata, dict):
+                metadata = {}
+            return content, metadata
+        except yaml.YAMLError as e:
+            logger.warning("Failed to parse YAML frontmatter: %s", e)
+            return content, None
+
+    def _append_section(self, title: str, content: str) -> None:
+        """Append a section to the prompt parts.
+
+        Args:
+            title: Section title
+            content: Section content
+        """
+        content = content.strip()
+        if not content:
+            return
+
+        if self._prompt_parts:
+            self._prompt_parts.append("")
+
+        self._prompt_parts.append(f"## {title}")
+        self._prompt_parts.append("")
+        self._prompt_parts.append(content)
+
+    def build(self) -> str:
+        """Build the system prompt from markdown files.
+
+        Returns:
+            Constructed system prompt string
+        """
+        self._prompt_parts = []
+        self._metadata = {}
+        self.loaded_count = 0
+
+        for filename in self.enabled_files:
+            content, metadata = self._load_file(filename)
+
+            if metadata:
+                self._metadata[filename] = metadata
+
+            if content:
+                title = self.TITLE_MAP.get(filename, filename.replace(".md", ""))
+                self._append_section(title, content)
+
+        if not self._prompt_parts:
+            logger.warning("No content loaded from workspace: %s", self.workspace_dir)
+            return DEFAULT_SYS_PROMPT
+
+        final_prompt = "\n".join(self._prompt_parts)
+
+        logger.debug(
+            "System prompt built from %d file(s), total length: %d chars",
+            self.loaded_count,
+            len(final_prompt),
+        )
+
+        return final_prompt
+
+    def get_metadata(self) -> Dict[str, Any]:
+        """Get metadata collected from YAML frontmatter.
+
+        Returns:
+            Dictionary mapping filenames to their metadata
+        """
+        return self._metadata.copy()
+
+    def get_agent_identity(self) -> Optional[Dict[str, Any]]:
+        """Extract agent identity from PROFILE.md metadata.
+
+        Returns:
+            Identity dict with name, role, etc. or None
+        """
+        profile_meta = self._metadata.get("PROFILE.md", {})
+        if not profile_meta:
+            return None
+
+        return {
+            "name": profile_meta.get("name", "Unknown"),
+            "role": profile_meta.get("role", ""),
+            "expertise": profile_meta.get("expertise", []),
+            "style": profile_meta.get("style", ""),
+        }
+
+
+def build_system_prompt_from_workspace(
+    workspace_dir: Path,
+    enabled_files: Optional[List[str]] = None,
+    agent_id: Optional[str] = None,
+    extra_context: Optional[str] = None,
+) -> str:
+    """Build system prompt from workspace markdown files.
+
+    This is the main entry point for building system prompts from
+    agent workspace directories.
+
+    Args:
+        workspace_dir: Directory containing markdown configuration files
+        enabled_files: List of filenames to load (if None, uses defaults)
+        agent_id: Agent identifier to include in system prompt
+        extra_context: Additional context to append to the prompt
+
+    Returns:
+        Constructed system prompt string
+    """
+    builder = PromptBuilder(
+        workspace_dir=workspace_dir,
+        enabled_files=enabled_files,
+    )
+
+    prompt = builder.build()
+
+    # Add agent identity header if agent_id provided
+    if agent_id and agent_id != "default":
+        identity_header = (
+            f"# Agent Identity\n\n"
+            f"Your agent ID is `{agent_id}`. "
+            f"This is your unique identifier in the multi-agent system.\n\n"
+        )
+        prompt = identity_header + prompt
+
+    # Append extra context if provided
+    if extra_context:
+        prompt = prompt + "\n\n" + extra_context
+
+    return prompt
+
+
+def build_bootstrap_guidance(language: str = "zh") -> str:
+    """Build bootstrap guidance message for first-time setup.
+
+    Args:
+        language: Language code (zh/en)
+
+    Returns:
+        Formatted bootstrap guidance message
+    """
+    if language == "zh":
+        return (
+            "# 引导模式\n"
+            "\n"
+            "工作目录中存在 `BOOTSTRAP.md` — 首次设置。\n"
+            "\n"
+            "1. 阅读 BOOTSTRAP.md，友好地表示初次见面，"
+            "引导用户完成设置。\n"
+            "2. 按照 BOOTSTRAP.md 的指示，"
+            "帮助用户定义你的身份和偏好。\n"
+            "3. 按指南创建/更新必要文件"
+            "（PROFILE.md、MEMORY.md 等）。\n"
+            "4. 完成后删除 BOOTSTRAP.md。\n"
+            "\n"
+            "如果用户希望跳过，直接回答下面的问题即可。\n"
+            "\n"
+            "---\n"
+            "\n"
+        )
+
+    return (
+        "# BOOTSTRAP MODE\n"
+        "\n"
+        "`BOOTSTRAP.md` exists — first-time setup.\n"
+        "\n"
+        "1. Read BOOTSTRAP.md, greet the user, "
+        "and guide them through setup.\n"
+        "2. Follow BOOTSTRAP.md instructions "
+        "to define identity and preferences.\n"
+        "3. Create/update files "
+        "(PROFILE.md, MEMORY.md, etc.) as described.\n"
+        "4. Delete BOOTSTRAP.md when done.\n"
+        "\n"
+        "If the user wants to skip, answer their "
+        "question directly instead.\n"
+        "\n"
+        "---\n"
+        "\n"
+    )
+
+
+__all__ = [
+    "PromptBuilder",
+    "build_system_prompt_from_workspace",
+    "build_bootstrap_guidance",
+    "DEFAULT_SYS_PROMPT",
+]
--- a/backend/agents/prompts/risk_manager/system.md
+++ b/backend/agents/prompts/risk_manager/system.md
@@ -8,6 +8,8 @@
 5. 基于市场条件建议仓位限制

 你的决策流程：
+1. 优先使用可用的风险工具量化集中度、波动率和保证金压力
+2. 结合工具结果与当前市场上下文做判断
 3. 生成可操作的风险警告和仓位限制建议
 4. 为你的风险评估提供清晰的理由

--- a/backend/agents/registry.py
+++ b/backend/agents/registry.py
@@ -0,0 +1,284 @@
+# -*- coding: utf-8 -*-
+"""Agent Registry - In-memory registry for agent management."""
+
+from dataclasses import dataclass, field
+from typing import Any, Dict, List, Optional
+
+
+@dataclass
+class AgentInfo:
+    """Information about a registered agent."""
+
+    agent_id: str
+    agent_type: str
+    workspace_id: str
+    config_path: str
+    agent_dir: str
+    status: str = "inactive"  # inactive, active, error
+    metadata: Dict[str, Any] = field(default_factory=dict)
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Serialize to dictionary."""
+        return {
+            "agent_id": self.agent_id,
+            "agent_type": self.agent_type,
+            "workspace_id": self.workspace_id,
+            "config_path": self.config_path,
+            "agent_dir": self.agent_dir,
+            "status": self.status,
+            "metadata": self.metadata,
+        }
+
+
+class AgentRegistry:
+    """In-memory registry for agent instances."""
+
+    def __init__(self):
+        """Initialize the agent registry."""
+        # Dictionary mapping agent_id -> AgentInfo
+        self._agents: Dict[str, AgentInfo] = {}
+        # Index mapping workspace_id -> set of agent_ids
+        self._workspace_index: Dict[str, set] = {}
+
+    def register(
+        self,
+        agent_id: str,
+        agent_type: str,
+        workspace_id: str,
+        config_path: str,
+        agent_dir: str,
+        status: str = "inactive",
+        metadata: Optional[Dict[str, Any]] = None,
+    ) -> AgentInfo:
+        """Register an agent in the registry.
+
+        Args:
+            agent_id: Unique identifier for the agent
+            agent_type: Type of agent
+            workspace_id: ID of the workspace containing the agent
+            config_path: Path to agent configuration file
+            agent_dir: Path to agent directory
+            status: Initial status (default: inactive)
+            metadata: Optional metadata dictionary
+
+        Returns:
+            AgentInfo instance
+
+        Raises:
+            ValueError: If agent_id is already registered
+        """
+        if agent_id in self._agents:
+            raise ValueError(f"Agent '{agent_id}' is already registered")
+
+        agent_info = AgentInfo(
+            agent_id=agent_id,
+            agent_type=agent_type,
+            workspace_id=workspace_id,
+            config_path=config_path,
+            agent_dir=agent_dir,
+            status=status,
+            metadata=metadata or {},
+        )
+
+        self._agents[agent_id] = agent_info
+
+        # Update workspace index
+        if workspace_id not in self._workspace_index:
+            self._workspace_index[workspace_id] = set()
+        self._workspace_index[workspace_id].add(agent_id)
+
+        return agent_info
+
+    def unregister(self, agent_id: str) -> bool:
+        """Unregister an agent.
+
+        Args:
+            agent_id: ID of the agent to unregister
+
+        Returns:
+            True if unregistered, False if agent wasn't registered
+        """
+        if agent_id not in self._agents:
+            return False
+
+        agent_info = self._agents[agent_id]
+
+        # Remove from workspace index
+        workspace_id = agent_info.workspace_id
+        if workspace_id in self._workspace_index:
+            self._workspace_index[workspace_id].discard(agent_id)
+            if not self._workspace_index[workspace_id]:
+                del self._workspace_index[workspace_id]
+
+        # Remove from agents dict
+        del self._agents[agent_id]
+
+        return True
+
+    def get(self, agent_id: str) -> Optional[AgentInfo]:
+        """Get agent information by ID.
+
+        Args:
+            agent_id: ID of the agent
+
+        Returns:
+            AgentInfo if found, None otherwise
+        """
+        return self._agents.get(agent_id)
+
+    def list_all(
+        self,
+        workspace_id: Optional[str] = None,
+        agent_type: Optional[str] = None,
+        status: Optional[str] = None,
+    ) -> List[AgentInfo]:
+        """List all registered agents with optional filtering.
+
+        Args:
+            workspace_id: Filter by workspace ID
+            agent_type: Filter by agent type
+            status: Filter by status
+
+        Returns:
+            List of AgentInfo instances
+        """
+        agents = list(self._agents.values())
+
+        if workspace_id:
+            agent_ids = self._workspace_index.get(workspace_id, set())
+            agents = [a for a in agents if a.agent_id in agent_ids]
+
+        if agent_type:
+            agents = [a for a in agents if a.agent_type == agent_type]
+
+        if status:
+            agents = [a for a in agents if a.status == status]
+
+        return agents
+
+    def update_status(self, agent_id: str, status: str) -> bool:
+        """Update the status of an agent.
+
+        Args:
+            agent_id: ID of the agent
+            status: New status value
+
+        Returns:
+            True if updated, False if agent not found
+        """
+        if agent_id not in self._agents:
+            return False
+
+        self._agents[agent_id].status = status
+        return True
+
+    def update_metadata(self, agent_id: str, metadata: Dict[str, Any]) -> bool:
+        """Update the metadata of an agent.
+
+        Args:
+            agent_id: ID of the agent
+            metadata: Metadata dictionary to merge
+
+        Returns:
+            True if updated, False if agent not found
+        """
+        if agent_id not in self._agents:
+            return False
+
+        self._agents[agent_id].metadata.update(metadata)
+        return True
+
+    def is_registered(self, agent_id: str) -> bool:
+        """Check if an agent is registered.
+
+        Args:
+            agent_id: ID of the agent
+
+        Returns:
+            True if registered, False otherwise
+        """
+        return agent_id in self._agents
+
+    def get_workspace_agents(self, workspace_id: str) -> List[AgentInfo]:
+        """Get all agents in a workspace.
+
+        Args:
+            workspace_id: ID of the workspace
+
+        Returns:
+            List of AgentInfo instances
+        """
+        agent_ids = self._workspace_index.get(workspace_id, set())
+        return [self._agents[agent_id] for agent_id in agent_ids if agent_id in self._agents]
+
+    def get_agent_count(self, workspace_id: Optional[str] = None) -> int:
+        """Get the count of registered agents.
+
+        Args:
+            workspace_id: Optional workspace ID to filter by
+
+        Returns:
+            Number of agents
+        """
+        if workspace_id:
+            return len(self._workspace_index.get(workspace_id, set()))
+        return len(self._agents)
+
+    def clear(self) -> None:
+        """Clear all registered agents."""
+        self._agents.clear()
+        self._workspace_index.clear()
+
+    def get_stats(self) -> Dict[str, Any]:
+        """Get registry statistics.
+
+        Returns:
+            Dictionary with registry statistics
+        """
+        stats = {
+            "total_agents": len(self._agents),
+            "workspaces": len(self._workspace_index),
+            "agents_by_workspace": {
+                ws_id: len(agent_ids)
+                for ws_id, agent_ids in self._workspace_index.items()
+            },
+            "agents_by_type": {},
+            "agents_by_status": {},
+        }
+
+        for agent in self._agents.values():
+            # Count by type
+            agent_type = agent.agent_type
+            stats["agents_by_type"][agent_type] = (
+                stats["agents_by_type"].get(agent_type, 0) + 1
+            )
+
+            # Count by status
+            status = agent.status
+            stats["agents_by_status"][status] = (
+                stats["agents_by_status"].get(status, 0) + 1
+            )
+
+        return stats
+
+
+# Global registry instance
+_global_registry: Optional[AgentRegistry] = None
+
+
+def get_registry() -> AgentRegistry:
+    """Get the global agent registry instance.
+
+    Returns:
+        AgentRegistry instance
+    """
+    global _global_registry
+    if _global_registry is None:
+        _global_registry = AgentRegistry()
+    return _global_registry
+
+
+def reset_registry() -> None:
+    """Reset the global registry (useful for testing)."""
+    global _global_registry
+    _global_registry = None
--- a/backend/agents/risk_manager.py
+++ b/backend/agents/risk_manager.py
@@ -11,9 +11,7 @@ from agentscope.message import Msg
 from agentscope.tool import Toolkit

 from ..utils.progress import progress
-from .prompt_loader import PromptLoader
-
-_prompt_loader = PromptLoader()
+from .prompt_factory import build_agent_system_prompt, clear_prompt_factory_cache


 class RiskAgent(ReActAgent):
@@ -29,6 +27,7 @@ class RiskAgent(ReActAgent):
        name: str = "risk_manager",
        config: Optional[Dict[str, Any]] = None,
        long_term_memory: Optional[LongTermMemoryBase] = None,
+        toolkit: Optional[Toolkit] = None,
    ):
        """
        Initialize Risk Manager Agent
@@ -40,13 +39,15 @@ class RiskAgent(ReActAgent):
            config: Configuration dictionary
            long_term_memory: Optional ReMeTaskLongTermMemory instance
        """
-        self.config = config or {}
+        object.__setattr__(self, "config", config or {})
+        object.__setattr__(self, "agent_id", name)
+
+        if toolkit is None:
+            toolkit = Toolkit()
+        object.__setattr__(self, "toolkit", toolkit)

        sys_prompt = self._load_system_prompt()

-        # Create dedicated toolkit for this agent
-        toolkit = Toolkit()
-
        kwargs = {
            "name": name,
            "sys_prompt": sys_prompt,
@@ -64,9 +65,10 @@ class RiskAgent(ReActAgent):

    def _load_system_prompt(self) -> str:
        """Load system prompt for risk manager"""
-        return _prompt_loader.load_prompt(
-            "risk_manager",
-            "system",
+        return build_agent_system_prompt(
+            agent_id=self.agent_id,
+            config_name=self.config.get("config_name", "default"),
+            toolkit=self.toolkit,
        )

    async def reply(self, x: Msg = None) -> Msg:
@@ -86,3 +88,23 @@ class RiskAgent(ReActAgent):
        progress.update_status(self.name, None, "Risk assessment completed")

        return result
+
+    def reload_runtime_assets(self, active_skill_dirs: Optional[list] = None) -> None:
+        """Reload toolkit and system prompt from current run assets."""
+        from .toolkit_factory import create_agent_toolkit
+
+        clear_prompt_factory_cache()
+        self.toolkit = create_agent_toolkit(
+            self.agent_id,
+            self.config.get("config_name", "default"),
+            active_skill_dirs=active_skill_dirs,
+        )
+        self._apply_runtime_sys_prompt(self._load_system_prompt())
+
+    def _apply_runtime_sys_prompt(self, sys_prompt: str) -> None:
+        """Update the prompt used by future turns and the cached system msg."""
+        self._sys_prompt = sys_prompt
+        for msg, _marks in self.memory.content:
+            if getattr(msg, "role", None) == "system":
+                msg.content = sys_prompt
+                break
--- a/backend/agents/skill_loader.py
+++ b/backend/agents/skill_loader.py
@@ -0,0 +1,388 @@
+# -*- coding: utf-8 -*-
+"""Skill loader for loading and validating skills from directories.
+
+提供从目录加载技能、解析SKILL.md frontmatter、获取工具列表等功能。
+"""
+import logging
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Set
+
+import yaml
+
+from backend.agents.skill_metadata import SkillMetadata, parse_skill_metadata
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class SkillInfo:
+    """完整的技能信息"""
+    name: str
+    description: str
+    version: str
+    source: str
+    path: Path
+    metadata: SkillMetadata
+    tools: List[str] = field(default_factory=list)
+    scripts: List[str] = field(default_factory=list)
+    references: List[str] = field(default_factory=list)
+    content: str = ""
+
+
+def load_skill_from_dir(skill_dir: Path, source: str = "unknown") -> Optional[Dict[str, Any]]:
+    """从目录加载技能
+
+    Args:
+        skill_dir: 技能目录路径
+        source: 技能来源 (builtin/customized/local/installed/active)
+
+    Returns:
+        技能信息字典，加载失败返回None
+    """
+    if not skill_dir.exists() or not skill_dir.is_dir():
+        logger.warning(f"Skill directory does not exist: {skill_dir}")
+        return None
+
+    skill_md = skill_dir / "SKILL.md"
+    if not skill_md.exists():
+        logger.warning(f"SKILL.md not found in: {skill_dir}")
+        return None
+
+    try:
+        # 解析元数据
+        metadata = parse_skill_metadata(skill_dir, source=source)
+
+        # 读取完整内容
+        content = skill_md.read_text(encoding="utf-8")
+
+        # 提取body (去掉frontmatter)
+        body = content
+        if content.startswith("---"):
+            parts = content.split("---", 2)
+            if len(parts) >= 3:
+                body = parts[2].strip()
+
+        # 获取工具列表
+        tools = get_skill_tools(skill_dir)
+
+        # 获取脚本列表
+        scripts = _get_skill_scripts(skill_dir)
+
+        # 获取参考资料列表
+        references = _get_skill_references(skill_dir)
+
+        return {
+            "name": metadata.name,
+            "skill_name": metadata.skill_name,
+            "description": metadata.description,
+            "version": metadata.version,
+            "source": source,
+            "path": str(skill_dir),
+            "content": body,
+            "tools": tools,
+            "scripts": scripts,
+            "references": references,
+            "metadata": metadata,
+        }
+    except Exception as e:
+        logger.error(f"Failed to load skill from {skill_dir}: {e}")
+        return None
+
+
+def parse_skill_metadata(skill_dir: Path, source: str = "unknown") -> SkillMetadata:
+    """解析技能元数据 (兼容已有函数)
+
+    Args:
+        skill_dir: 技能目录路径
+        source: 技能来源
+
+    Returns:
+        SkillMetadata对象
+    """
+    from backend.agents.skill_metadata import parse_skill_metadata as _parse
+    return _parse(skill_dir, source=source)
+
+
+def get_skill_tools(skill_dir: Path) -> List[str]:
+    """获取技能提供的工具列表
+
+    从SKILL.md frontmatter的tools字段和scripts目录解析工具。
+
+    Args:
+        skill_dir: 技能目录路径
+
+    Returns:
+        工具名称列表
+    """
+    tools: Set[str] = set()
+
+    # 1. 从SKILL.md frontmatter读取tools字段
+    skill_md = skill_dir / "SKILL.md"
+    if skill_md.exists():
+        try:
+            raw = skill_md.read_text(encoding="utf-8").strip()
+            if raw.startswith("---"):
+                parts = raw.split("---", 2)
+                if len(parts) >= 3:
+                    try:
+                        frontmatter = yaml.safe_load(parts[1].strip()) or {}
+                        if isinstance(frontmatter, dict):
+                            tools_list = frontmatter.get("tools", [])
+                            if isinstance(tools_list, str):
+                                tools.add(tools_list.strip())
+                            elif isinstance(tools_list, list):
+                                for tool in tools_list:
+                                    if isinstance(tool, str):
+                                        tools.add(tool.strip())
+                    except yaml.YAMLError:
+                        pass
+        except Exception as e:
+            logger.warning(f"Failed to parse tools from SKILL.md: {e}")
+
+    # 2. 从scripts目录推断工具
+    scripts_dir = skill_dir / "scripts"
+    if scripts_dir.exists() and scripts_dir.is_dir():
+        for script in scripts_dir.iterdir():
+            if script.is_file() and not script.name.startswith("_"):
+                # 去掉扩展名作为工具名
+                tool_name = script.stem
+                tools.add(tool_name)
+
+    return sorted(list(tools))
+
+
+def _get_skill_scripts(skill_dir: Path) -> List[str]:
+    """获取技能脚本列表
+
+    Args:
+        skill_dir: 技能目录路径
+
+    Returns:
+        脚本相对路径列表 (相对于scripts目录)
+    """
+    scripts: List[str] = []
+    scripts_dir = skill_dir / "scripts"
+
+    if not scripts_dir.exists():
+        return scripts
+
+    try:
+        for item in scripts_dir.rglob("*"):
+            if item.is_file() and not item.name.startswith("_"):
+                rel_path = item.relative_to(scripts_dir)
+                scripts.append(str(rel_path))
+    except Exception as e:
+        logger.warning(f"Failed to list scripts in {skill_dir}: {e}")
+
+    return sorted(scripts)
+
+
+def _get_skill_references(skill_dir: Path) -> List[str]:
+    """获取技能参考资料列表
+
+    Args:
+        skill_dir: 技能目录路径
+
+    Returns:
+        参考资料相对路径列表 (相对于references目录)
+    """
+    refs: List[str] = []
+    refs_dir = skill_dir / "references"
+
+    if not refs_dir.exists():
+        return refs
+
+    try:
+        for item in refs_dir.rglob("*"):
+            if item.is_file():
+                rel_path = item.relative_to(refs_dir)
+                refs.append(str(rel_path))
+    except Exception as e:
+        logger.warning(f"Failed to list references in {skill_dir}: {e}")
+
+    return sorted(refs)
+
+
+def validate_skill(skill_dir: Path) -> Dict[str, Any]:
+    """验证技能格式
+
+    检查技能目录结构是否符合规范。
+
+    Args:
+        skill_dir: 技能目录路径
+
+    Returns:
+        验证结果字典，包含:
+        - valid: 是否有效
+        - errors: 错误列表
+        - warnings: 警告列表
+    """
+    errors: List[str] = []
+    warnings: List[str] = []
+
+    # 检查目录存在
+    if not skill_dir.exists():
+        errors.append(f"Skill directory does not exist: {skill_dir}")
+        return {"valid": False, "errors": errors, "warnings": warnings}
+
+    if not skill_dir.is_dir():
+        errors.append(f"Path is not a directory: {skill_dir}")
+        return {"valid": False, "errors": errors, "warnings": warnings}
+
+    # 检查SKILL.md
+    skill_md = skill_dir / "SKILL.md"
+    if not skill_md.exists():
+        errors.append("SKILL.md is required but not found")
+        return {"valid": False, "errors": errors, "warnings": warnings}
+
+    # 解析frontmatter
+    try:
+        content = skill_md.read_text(encoding="utf-8").strip()
+        if not content.startswith("---"):
+            warnings.append("SKILL.md should have YAML frontmatter (starts with ---)")
+        else:
+            parts = content.split("---", 2)
+            if len(parts) < 3:
+                errors.append("Invalid YAML frontmatter format")
+            else:
+                try:
+                    frontmatter = yaml.safe_load(parts[1].strip()) or {}
+                    if not isinstance(frontmatter, dict):
+                        errors.append("YAML frontmatter must be a dictionary")
+                    else:
+                        # 检查必需字段
+                        if "name" not in frontmatter:
+                            warnings.append("Frontmatter should have 'name' field")
+                        if "description" not in frontmatter:
+                            warnings.append("Frontmatter should have 'description' field")
+
+                        # 检查version字段
+                        version = frontmatter.get("version")
+                        if version and not isinstance(version, str):
+                            warnings.append("'version' should be a string")
+
+                        # 检查tools字段
+                        tools = frontmatter.get("tools")
+                        if tools and not isinstance(tools, (str, list)):
+                            warnings.append("'tools' should be a string or list")
+
+                except yaml.YAMLError as e:
+                    errors.append(f"Invalid YAML in frontmatter: {e}")
+    except Exception as e:
+        errors.append(f"Failed to read SKILL.md: {e}")
+
+    # 检查body内容
+    try:
+        content = skill_md.read_text(encoding="utf-8")
+        body = content
+        if content.startswith("---"):
+            parts = content.split("---", 2)
+            if len(parts) >= 3:
+                body = parts[2].strip()
+
+        if not body:
+            warnings.append("SKILL.md body is empty")
+        elif len(body) < 50:
+            warnings.append("SKILL.md body is very short, consider adding more details")
+    except Exception as e:
+        errors.append(f"Failed to validate body: {e}")
+
+    # 检查scripts目录
+    scripts_dir = skill_dir / "scripts"
+    if scripts_dir.exists():
+        if not scripts_dir.is_dir():
+            errors.append("'scripts' exists but is not a directory")
+        else:
+            # 检查是否有可执行脚本
+            has_scripts = any(
+                f.is_file() and not f.name.startswith("_")
+                for f in scripts_dir.iterdir()
+            )
+            if not has_scripts:
+                warnings.append("scripts directory exists but contains no valid scripts")
+
+    # 检查references目录
+    refs_dir = skill_dir / "references"
+    if refs_dir.exists() and not refs_dir.is_dir():
+        errors.append("'references' exists but is not a directory")
+
+    return {
+        "valid": len(errors) == 0,
+        "errors": errors,
+        "warnings": warnings,
+    }
+
+
+def load_skills_from_directory(
+    directory: Path,
+    source: str = "unknown",
+    recursive: bool = False,
+) -> List[Dict[str, Any]]:
+    """从目录加载所有技能
+
+    Args:
+        directory: 包含技能目录的父目录
+        source: 技能来源标识
+        recursive: 是否递归搜索子目录
+
+    Returns:
+        技能信息列表
+    """
+    skills: List[Dict[str, Any]] = []
+
+    if not directory.exists() or not directory.is_dir():
+        logger.warning(f"Directory does not exist: {directory}")
+        return skills
+
+    try:
+        for item in directory.iterdir():
+            if not item.is_dir():
+                continue
+
+            # 检查是否是技能目录 (包含SKILL.md)
+            if (item / "SKILL.md").exists():
+                skill_info = load_skill_from_dir(item, source=source)
+                if skill_info:
+                    skills.append(skill_info)
+            elif recursive:
+                # 递归搜索子目录
+                sub_skills = load_skills_from_directory(item, source, recursive)
+                skills.extend(sub_skills)
+
+    except Exception as e:
+        logger.error(f"Failed to load skills from {directory}: {e}")
+
+    return skills
+
+
+def get_skill_manifest(skill_dir: Path) -> Dict[str, Any]:
+    """获取技能清单
+
+    生成技能的详细清单，用于调试和展示。
+
+    Args:
+        skill_dir: 技能目录路径
+
+    Returns:
+        技能清单字典
+    """
+    info = load_skill_from_dir(skill_dir)
+    if not info:
+        return {"error": "Failed to load skill"}
+
+    validation = validate_skill(skill_dir)
+
+    return {
+        "name": info["name"],
+        "skill_name": info["skill_name"],
+        "version": info["version"],
+        "description": info["description"],
+        "source": info["source"],
+        "path": info["path"],
+        "tools": info["tools"],
+        "scripts": info["scripts"],
+        "references": info["references"],
+        "validation": validation,
+        "content_preview": info["content"][:500] + "..." if len(info["content"]) > 500 else info["content"],
+    }
--- a/backend/agents/skill_metadata.py
+++ b/backend/agents/skill_metadata.py
@@ -0,0 +1,83 @@
+# -*- coding: utf-8 -*-
+"""Skill metadata parsing helpers for SKILL.md files."""
+
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import List
+
+import yaml
+
+
+@dataclass(frozen=True)
+class SkillMetadata:
+    """Parsed metadata for a skill package."""
+
+    skill_name: str
+    path: Path
+    source: str
+    name: str
+    description: str
+    version: str = ""
+    tools: List[str] = field(default_factory=list)
+    allowed_tools: List[str] = field(default_factory=list)
+    denied_tools: List[str] = field(default_factory=list)
+
+
+def parse_skill_metadata(skill_dir: Path, source: str) -> SkillMetadata:
+    """Parse SKILL.md frontmatter with a forgiving schema."""
+    skill_name = skill_dir.name
+    skill_file = skill_dir / "SKILL.md"
+    if not skill_file.exists():
+        return SkillMetadata(
+            skill_name=skill_name,
+            path=skill_dir,
+            source=source,
+            name=skill_name,
+            description="",
+        )
+
+    raw = skill_file.read_text(encoding="utf-8").strip()
+    frontmatter = {}
+    body = raw
+    if raw.startswith("---"):
+        parts = raw.split("---", 2)
+        if len(parts) >= 3:
+            try:
+                frontmatter = yaml.safe_load(parts[1].strip()) or {}
+            except yaml.YAMLError:
+                frontmatter = {}
+            body = parts[2].strip()
+    if not isinstance(frontmatter, dict):
+        frontmatter = {}
+
+    description = str(frontmatter.get("description") or "").strip()
+    if not description and body:
+        description = body.splitlines()[0].strip().lstrip("#").strip()
+
+    return SkillMetadata(
+        skill_name=skill_name,
+        path=skill_dir,
+        source=source,
+        name=str(frontmatter.get("name") or skill_name).strip() or skill_name,
+        description=description,
+        version=str(frontmatter.get("version") or "").strip(),
+        tools=_string_list(frontmatter.get("tools")),
+        allowed_tools=_string_list(frontmatter.get("allowed_tools")),
+        denied_tools=_string_list(frontmatter.get("denied_tools")),
+    )
+
+
+def _string_list(value) -> List[str]:
+    if isinstance(value, str):
+        item = value.strip()
+        return [item] if item else []
+    if not isinstance(value, list):
+        return []
+    seen: List[str] = []
+    for item in value:
+        if not isinstance(item, str):
+            continue
+        normalized = item.strip()
+        if normalized and normalized not in seen:
+            seen.append(normalized)
+    return seen
--- a/backend/agents/skills_manager.py
+++ b/backend/agents/skills_manager.py
@@ -0,0 +1,875 @@
+# -*- coding: utf-8 -*-
+"""Manage agent-installed and run-active skill directories for each run."""
+
+from pathlib import Path
+import shutil
+import tempfile
+import zipfile
+from threading import Lock
+from typing import Any, Dict, Iterable, Iterator, List, Optional, Set
+from urllib.parse import urlparse
+from urllib.request import urlretrieve
+
+import yaml
+
+from backend.agents.agent_workspace import load_agent_workspace_config
+from backend.agents.skill_metadata import SkillMetadata, parse_skill_metadata
+from backend.agents.skill_loader import validate_skill
+from backend.config.bootstrap_config import get_bootstrap_config_for_run
+
+try:
+    from watchdog.observers import Observer
+    from watchdog.events import FileSystemEventHandler, FileSystemEvent
+    WATCHDOG_AVAILABLE = True
+except ImportError:
+    WATCHDOG_AVAILABLE = False
+    Observer = None
+    FileSystemEventHandler = object
+    FileSystemEvent = object  # type: ignore[misc,assignment]
+
+
+class SkillsManager:
+    """Sync named skills into a run-scoped active skills workspace."""
+
+    def __init__(self, project_root: Path | None = None):
+        self.project_root = (
+            project_root or Path(__file__).resolve().parents[2]
+        )
+        self.builtin_root = self.project_root / "backend" / "skills" / "builtin"
+        self.customized_root = (
+            self.project_root / "backend" / "skills" / "customized"
+        )
+        self.runs_root = self.project_root / "runs"
+        self._lock = Lock()
+
+    def get_active_root(self, config_name: str) -> Path:
+        return self.runs_root / config_name / "skills" / "active"
+
+    def get_agent_skills_root(self, config_name: str, agent_id: str) -> Path:
+        return self.get_agent_asset_dir(config_name, agent_id) / "skills"
+
+    def get_agent_active_root(self, config_name: str, agent_id: str) -> Path:
+        return self.get_agent_skills_root(config_name, agent_id) / "active"
+
+    def get_agent_installed_root(self, config_name: str, agent_id: str) -> Path:
+        return self.get_agent_skills_root(config_name, agent_id) / "installed"
+
+    def get_agent_disabled_root(self, config_name: str, agent_id: str) -> Path:
+        return self.get_agent_skills_root(config_name, agent_id) / "disabled"
+
+    def get_agent_local_root(self, config_name: str, agent_id: str) -> Path:
+        return self.get_agent_skills_root(config_name, agent_id) / "local"
+
+    def get_activation_manifest_path(self, config_name: str) -> Path:
+        return self.runs_root / config_name / "skills" / "activation.yaml"
+
+    def get_agent_asset_dir(self, config_name: str, agent_id: str) -> Path:
+        return self.runs_root / config_name / "agents" / agent_id
+
+    def list_skill_catalog(self) -> List[SkillMetadata]:
+        """Return builtin/customized skills with parsed metadata."""
+        catalog: Dict[str, SkillMetadata] = {}
+
+        for source, root in (
+            ("builtin", self.builtin_root),
+            ("customized", self.customized_root),
+        ):
+            if not root.exists():
+                continue
+            for skill_dir in sorted(root.iterdir(), key=lambda item: item.name):
+                if not skill_dir.is_dir():
+                    continue
+                if not (skill_dir / "SKILL.md").exists():
+                    continue
+                metadata = parse_skill_metadata(skill_dir, source=source)
+                catalog[metadata.skill_name] = metadata
+
+        return sorted(catalog.values(), key=lambda item: item.skill_name)
+
+    def list_agent_skill_catalog(
+        self,
+        config_name: str,
+        agent_id: str,
+    ) -> List[SkillMetadata]:
+        """Return shared plus agent-local skills for one agent."""
+        catalog = {
+            item.skill_name: item
+            for item in self.list_skill_catalog()
+        }
+        for item in self.list_agent_local_skills(config_name, agent_id):
+            catalog[item.skill_name] = item
+        return sorted(catalog.values(), key=lambda item: item.skill_name)
+
+    def list_active_skill_metadata(
+        self,
+        config_name: str,
+        agent_id: str,
+    ) -> List[SkillMetadata]:
+        """Return metadata for active skills synced for one agent."""
+        active_root = self.get_agent_active_root(config_name, agent_id)
+        if not active_root.exists():
+            return []
+
+        items: List[SkillMetadata] = []
+        for skill_dir in sorted(active_root.iterdir(), key=lambda item: item.name):
+            if not skill_dir.is_dir():
+                continue
+            if not (skill_dir / "SKILL.md").exists():
+                continue
+            items.append(parse_skill_metadata(skill_dir, source="active"))
+        return items
+
+    def list_agent_local_skills(
+        self,
+        config_name: str,
+        agent_id: str,
+    ) -> List[SkillMetadata]:
+        """Return metadata for agent-private local skills."""
+        local_root = self.get_agent_local_root(config_name, agent_id)
+        if not local_root.exists():
+            return []
+
+        items: List[SkillMetadata] = []
+        for skill_dir in sorted(local_root.iterdir(), key=lambda item: item.name):
+            if not skill_dir.is_dir():
+                continue
+            if not (skill_dir / "SKILL.md").exists():
+                continue
+            items.append(parse_skill_metadata(skill_dir, source="local"))
+        return items
+
+    def load_skill_document(self, skill_name: str) -> Dict[str, object]:
+        """Return skill metadata plus markdown body for one skill."""
+        source_dir = self._resolve_source_dir(skill_name)
+        return self._load_skill_document_from_dir(
+            source_dir,
+            source="customized" if source_dir.parent == self.customized_root else "builtin",
+        )
+
+    def load_agent_skill_document(
+        self,
+        config_name: str,
+        agent_id: str,
+        skill_name: str,
+    ) -> Dict[str, object]:
+        """Return skill metadata plus markdown body for one agent-visible skill."""
+        source_dir = self._resolve_agent_skill_source_dir(
+            config_name=config_name,
+            agent_id=agent_id,
+            skill_name=skill_name,
+        )
+        source = "local"
+        if source_dir.parent == self.customized_root:
+            source = "customized"
+        elif source_dir.parent == self.builtin_root:
+            source = "builtin"
+        elif source_dir.parent == self.get_agent_installed_root(config_name, agent_id):
+            source = "installed"
+        return self._load_skill_document_from_dir(source_dir, source=source)
+
+    def create_agent_local_skill(
+        self,
+        config_name: str,
+        agent_id: str,
+        skill_name: str,
+    ) -> Path:
+        """Create a new local skill directory with a default SKILL.md."""
+        normalized = _normalize_skill_name(skill_name)
+        if not normalized:
+            raise ValueError("Skill name is required.")
+        local_root = self.get_agent_local_root(config_name, agent_id)
+        local_root.mkdir(parents=True, exist_ok=True)
+        skill_dir = local_root / normalized
+        if skill_dir.exists():
+            raise FileExistsError(f"Local skill already exists: {normalized}")
+        skill_dir.mkdir(parents=True, exist_ok=False)
+        (skill_dir / "SKILL.md").write_text(
+            "---\n"
+            f"name: {normalized}\n"
+            "description: 当用户提出与该本地技能相关的专门任务时，应使用此技能。\n"
+            "version: 1.0.0\n"
+            "---\n\n"
+            f"# {normalized}\n\n"
+            "在这里描述该交易员的专有分析流程、判断框架和可复用步骤。\n",
+            encoding="utf-8",
+        )
+        return skill_dir
+
+    def install_external_skill_for_agent(
+        self,
+        config_name: str,
+        agent_id: str,
+        source: str,
+        *,
+        skill_name: str | None = None,
+        activate: bool = True,
+    ) -> Dict[str, object]:
+        """
+        Install an external skill into one agent's local skill space.
+
+        Supports:
+        - local skill directory containing SKILL.md
+        - local zip archive containing one skill directory
+        - http(s) URL to zip archive
+        """
+        source_path = self._resolve_external_source_path(source)
+        skill_dir = self._resolve_external_skill_dir(source_path)
+        metadata = parse_skill_metadata(skill_dir, source="external")
+        final_name = _normalize_skill_name(skill_name or metadata.skill_name or skill_dir.name)
+        if not final_name:
+            raise ValueError("Could not determine skill name from external source.")
+
+        target_dir = self.get_agent_local_root(config_name, agent_id) / final_name
+        target_dir.parent.mkdir(parents=True, exist_ok=True)
+        if target_dir.exists():
+            shutil.rmtree(target_dir)
+        shutil.copytree(skill_dir, target_dir)
+
+        validation = validate_skill(target_dir)
+        if not validation.get("valid", False):
+            shutil.rmtree(target_dir, ignore_errors=True)
+            raise ValueError(
+                "Installed skill is invalid: "
+                + "; ".join(validation.get("errors", []))
+            )
+
+        if activate:
+            self.update_agent_skill_overrides(
+                config_name=config_name,
+                agent_id=agent_id,
+                enable=[final_name],
+            )
+        return {
+            "skill_name": final_name,
+            "target_dir": str(target_dir),
+            "activated": activate,
+            "warnings": validation.get("warnings", []),
+        }
+
+    def update_agent_local_skill(
+        self,
+        config_name: str,
+        agent_id: str,
+        skill_name: str,
+        content: str,
+    ) -> Path:
+        """Overwrite one agent-local SKILL.md."""
+        normalized = _normalize_skill_name(skill_name)
+        if not normalized:
+            raise ValueError("Skill name is required.")
+        skill_dir = self.get_agent_local_root(config_name, agent_id) / normalized
+        if not skill_dir.exists():
+            raise FileNotFoundError(f"Unknown local skill: {normalized}")
+        (skill_dir / "SKILL.md").write_text(content, encoding="utf-8")
+        return skill_dir
+
+    def delete_agent_local_skill(
+        self,
+        config_name: str,
+        agent_id: str,
+        skill_name: str,
+    ) -> None:
+        """Delete one agent-local skill directory."""
+        normalized = _normalize_skill_name(skill_name)
+        if not normalized:
+            raise ValueError("Skill name is required.")
+        skill_dir = self.get_agent_local_root(config_name, agent_id) / normalized
+        if not skill_dir.exists():
+            raise FileNotFoundError(f"Unknown local skill: {normalized}")
+        shutil.rmtree(skill_dir)
+
+    def _load_skill_document_from_dir(
+        self,
+        source_dir: Path,
+        *,
+        source: str,
+    ) -> Dict[str, object]:
+        """Return metadata plus markdown body for one resolved skill directory."""
+        metadata = parse_skill_metadata(
+            source_dir,
+            source=source,
+        )
+        skill_file = source_dir / "SKILL.md"
+        raw = skill_file.read_text(encoding="utf-8").strip() if skill_file.exists() else ""
+        body = raw
+        if raw.startswith("---"):
+            parts = raw.split("---", 2)
+            if len(parts) >= 3:
+                body = parts[2].strip()
+
+        return {
+            "skill_name": metadata.skill_name,
+            "name": metadata.name,
+            "description": metadata.description,
+            "version": metadata.version,
+            "tools": metadata.tools,
+            "source": metadata.source,
+            "content": body,
+        }
+
+    def _resolve_external_source_path(self, source: str) -> Path:
+        """Resolve source into a local path; download URL when needed."""
+        parsed = urlparse(source)
+        if parsed.scheme in {"http", "https"}:
+            suffix = Path(parsed.path).suffix or ".zip"
+            with tempfile.NamedTemporaryFile(suffix=suffix, delete=False) as tmp:
+                temp_path = Path(tmp.name)
+            urlretrieve(source, temp_path)
+            return temp_path
+        return Path(source).expanduser().resolve()
+
+    def _resolve_external_skill_dir(self, source_path: Path) -> Path:
+        """Resolve external source path to a skill directory containing SKILL.md."""
+        if not source_path.exists():
+            raise FileNotFoundError(f"Source does not exist: {source_path}")
+
+        if source_path.is_dir():
+            if (source_path / "SKILL.md").exists():
+                return source_path
+            children = [
+                item for item in source_path.iterdir()
+                if item.is_dir() and (item / "SKILL.md").exists()
+            ]
+            if len(children) == 1:
+                return children[0]
+            raise ValueError(
+                "Source directory must contain SKILL.md "
+                "or exactly one child directory containing SKILL.md."
+            )
+
+        if source_path.suffix.lower() != ".zip":
+            raise ValueError("External source file must be a .zip archive.")
+
+        temp_root = Path(tempfile.mkdtemp(prefix="external_skill_"))
+        with zipfile.ZipFile(source_path, "r") as archive:
+            archive.extractall(temp_root)
+
+        candidates = [
+            item.parent
+            for item in temp_root.rglob("SKILL.md")
+            if item.is_file()
+        ]
+        unique = []
+        for item in candidates:
+            if item not in unique:
+                unique.append(item)
+        if len(unique) != 1:
+            raise ValueError(
+                "Zip archive must contain exactly one skill directory with SKILL.md."
+            )
+        return unique[0]
+
+    def update_agent_skill_overrides(
+        self,
+        config_name: str,
+        agent_id: str,
+        *,
+        enable: Iterable[str] | None = None,
+        disable: Iterable[str] | None = None,
+    ) -> Dict[str, List[str]]:
+        """Persist per-agent enabled/disabled skill overrides in agent.yaml."""
+        asset_dir = self.get_agent_asset_dir(config_name, agent_id)
+        asset_dir.mkdir(parents=True, exist_ok=True)
+        config_path = asset_dir / "agent.yaml"
+        current = load_agent_workspace_config(config_path)
+        values = dict(current.values)
+
+        enabled = _dedupe_preserve_order(current.enabled_skills)
+        disabled_set = set(current.disabled_skills)
+
+        for skill_name in enable or []:
+            if skill_name not in enabled:
+                enabled.append(skill_name)
+            disabled_set.discard(skill_name)
+
+        for skill_name in disable or []:
+            disabled_set.add(skill_name)
+            enabled = [item for item in enabled if item != skill_name]
+
+        values["enabled_skills"] = enabled
+        values["disabled_skills"] = sorted(disabled_set)
+        config_path.write_text(
+            yaml.safe_dump(values, allow_unicode=True, sort_keys=False),
+            encoding="utf-8",
+        )
+        return {
+            "enabled_skills": enabled,
+            "disabled_skills": sorted(disabled_set),
+        }
+
+    def forget_agent_skill_overrides(
+        self,
+        config_name: str,
+        agent_id: str,
+        skill_names: Iterable[str],
+    ) -> Dict[str, List[str]]:
+        """Remove skills from both enabled/disabled overrides in agent.yaml."""
+        asset_dir = self.get_agent_asset_dir(config_name, agent_id)
+        asset_dir.mkdir(parents=True, exist_ok=True)
+        config_path = asset_dir / "agent.yaml"
+        current = load_agent_workspace_config(config_path)
+        values = dict(current.values)
+        removed = set(skill_names)
+
+        enabled = [item for item in current.enabled_skills if item not in removed]
+        disabled = [item for item in current.disabled_skills if item not in removed]
+
+        values["enabled_skills"] = enabled
+        values["disabled_skills"] = disabled
+        config_path.write_text(
+            yaml.safe_dump(values, allow_unicode=True, sort_keys=False),
+            encoding="utf-8",
+        )
+        return {
+            "enabled_skills": enabled,
+            "disabled_skills": disabled,
+        }
+
+    def ensure_activation_manifest(self, config_name: str) -> Path:
+        manifest_path = self.get_activation_manifest_path(config_name)
+        manifest_path.parent.mkdir(parents=True, exist_ok=True)
+        if not manifest_path.exists():
+            manifest_path.write_text(
+                "global_enabled_skills: []\n"
+                "global_disabled_skills: []\n"
+                "agent_enabled_skills: {}\n"
+                "agent_disabled_skills: {}\n",
+                encoding="utf-8",
+            )
+        return manifest_path
+
+    def load_activation_manifest(self, config_name: str) -> Dict[str, object]:
+        manifest_path = self.ensure_activation_manifest(config_name)
+        with open(manifest_path, "r", encoding="utf-8") as file:
+            parsed = yaml.safe_load(file) or {}
+        return parsed if isinstance(parsed, dict) else {}
+
+    def _resolve_source_dir(self, skill_name: str) -> Path:
+        customized_dir = self.customized_root / skill_name
+        if customized_dir.exists():
+            return customized_dir
+
+        builtin_dir = self.builtin_root / skill_name
+        if builtin_dir.exists():
+            return builtin_dir
+
+        raise FileNotFoundError(f"Unknown skill: {skill_name}")
+
+    def _resolve_agent_skill_source_dir(
+        self,
+        config_name: str,
+        agent_id: str,
+        skill_name: str,
+    ) -> Path:
+        """Resolve one skill from the agent-local workspace or shared registry."""
+        for root in (
+            self.get_agent_local_root(config_name, agent_id),
+            self.get_agent_installed_root(config_name, agent_id),
+        ):
+            candidate = root / skill_name
+            if candidate.exists() and (candidate / "SKILL.md").exists():
+                return candidate
+        return self._resolve_source_dir(skill_name)
+
+    def _skill_exists_for_agent(
+        self,
+        config_name: str,
+        agent_id: str,
+        skill_name: str,
+    ) -> bool:
+        try:
+            self._resolve_agent_skill_source_dir(config_name, agent_id, skill_name)
+        except FileNotFoundError:
+            return False
+        return True
+
+    def _persist_runtime_edits(
+        self,
+        config_name: str,
+        skill_name: str,
+        active_dir: Path,
+    ) -> None:
+        """
+        Persist run-time edits from active skills into customized skills.
+
+        This keeps active skill experiments from being lost on the next reload
+        while still allowing the active directory to be re-synced cleanly.
+        """
+        if not active_dir.exists():
+            return
+
+        source_dir = self._resolve_source_dir(skill_name)
+        if active_dir.resolve() == source_dir.resolve():
+            return
+
+        if not self._directories_match(active_dir, source_dir):
+            customized_dir = self.customized_root / skill_name
+            customized_dir.parent.mkdir(parents=True, exist_ok=True)
+            if customized_dir.exists():
+                shutil.rmtree(customized_dir)
+            shutil.copytree(active_dir, customized_dir)
+
+    @staticmethod
+    def _directories_match(left: Path, right: Path) -> bool:
+        """Compare two directory trees by file contents."""
+        if not left.exists() or not right.exists():
+            return False
+
+        left_items = sorted(
+            path.relative_to(left)
+            for path in left.rglob("*")
+        )
+        right_items = sorted(
+            path.relative_to(right)
+            for path in right.rglob("*")
+        )
+        if left_items != right_items:
+            return False
+
+        for relative_path in left_items:
+            left_path = left / relative_path
+            right_path = right / relative_path
+            if left_path.is_dir() != right_path.is_dir():
+                return False
+            if left_path.is_file():
+                if left_path.read_bytes() != right_path.read_bytes():
+                    return False
+        return True
+
+    def resolve_agent_skill_names(
+        self,
+        config_name: str,
+        agent_id: str,
+        default_skills: Iterable[str],
+    ) -> List[str]:
+        """Resolve final skill names after bootstrap and activation overlays."""
+        bootstrap = get_bootstrap_config_for_run(self.project_root, config_name)
+        override = bootstrap.agent_override(agent_id)
+        skills = list(override.get("skills", list(default_skills)))
+        agent_config = load_agent_workspace_config(
+            self.get_agent_asset_dir(config_name, agent_id) / "agent.yaml",
+        )
+
+        for skill_name in agent_config.enabled_skills:
+            if skill_name not in skills:
+                skills.append(skill_name)
+
+        manifest = self.load_activation_manifest(config_name)
+        for skill_name in manifest.get("global_enabled_skills", []):
+            if skill_name not in skills:
+                skills.append(skill_name)
+
+        for skill_name in manifest.get("agent_enabled_skills", {}).get(agent_id, []):
+            if skill_name not in skills:
+                skills.append(skill_name)
+
+        disabled = set(manifest.get("global_disabled_skills", []))
+        disabled.update(
+            manifest.get("agent_disabled_skills", {}).get(agent_id, []),
+        )
+        disabled.update(agent_config.disabled_skills)
+
+        for item in self.list_agent_local_skills(config_name, agent_id):
+            if item.skill_name not in skills:
+                skills.append(item.skill_name)
+
+        return [
+            skill
+            for skill in skills
+            if skill not in disabled
+            and self._skill_exists_for_agent(config_name, agent_id, skill)
+        ]
+
+    def sync_skill_dirs(
+        self,
+        target_root: Path,
+        skill_sources: Dict[str, Path],
+    ) -> List[Path]:
+        """Sync selected skill directories into one target root."""
+        target_root.mkdir(parents=True, exist_ok=True)
+
+        synced_paths: List[Path] = []
+        wanted = set(skill_sources)
+
+        for existing in target_root.iterdir():
+            if existing.is_dir() and existing.name not in wanted:
+                shutil.rmtree(existing)
+
+        for skill_name, source_dir in skill_sources.items():
+            target_dir = target_root / skill_name
+            if target_dir.exists():
+                shutil.rmtree(target_dir)
+            shutil.copytree(source_dir, target_dir)
+            synced_paths.append(target_dir)
+
+        return synced_paths
+
+    def sync_active_skills(
+        self,
+        target_root: Path,
+        skill_names: Iterable[str],
+    ) -> List[Path]:
+        """Sync selected shared skills into one active directory."""
+        skill_sources = {
+            skill_name: self._resolve_source_dir(skill_name)
+            for skill_name in skill_names
+        }
+        return self.sync_skill_dirs(target_root, skill_sources)
+
+    def prepare_active_skills(
+        self,
+        config_name: str,
+        agent_defaults: Dict[str, Iterable[str]],
+        auto_reload: bool = False,
+    ) -> Dict[str, List[Path]]:
+        """Resolve all agent skills into per-agent installed/active workspaces."""
+        resolved: Dict[str, List[str]] = {}
+        union: List[str] = []
+
+        for agent_id, default_skills in agent_defaults.items():
+            resolved_skills = self.resolve_agent_skill_names(
+                config_name=config_name,
+                agent_id=agent_id,
+                default_skills=default_skills,
+            )
+            resolved[agent_id] = resolved_skills
+            for skill_name in resolved_skills:
+                if skill_name not in union:
+                    union.append(skill_name)
+
+        # Maintain the legacy union directory for compatibility/debugging.
+        # Agent-local skills remain private to the agent workspace.
+        self.sync_active_skills(
+            target_root=self.get_active_root(config_name),
+            skill_names=[
+                skill_name
+                for skill_name in union
+                if self._is_shared_skill(skill_name)
+            ],
+        )
+
+        active_map: Dict[str, List[Path]] = {}
+        for agent_id, skill_names in resolved.items():
+            installed_sources = {
+                skill_name: self._resolve_source_dir(skill_name)
+                for skill_name in skill_names
+                if (self.get_agent_local_root(config_name, agent_id) / skill_name).exists() is False
+            }
+            installed_paths = self.sync_skill_dirs(
+                target_root=self.get_agent_installed_root(config_name, agent_id),
+                skill_sources=installed_sources,
+            )
+
+            local_root = self.get_agent_local_root(config_name, agent_id)
+            local_sources = {
+                skill_name: local_root / skill_name
+                for skill_name in skill_names
+                if (local_root / skill_name).exists()
+            }
+            active_sources = {
+                path.name: path for path in installed_paths
+            }
+            active_sources.update(local_sources)
+            active_map[agent_id] = self.sync_skill_dirs(
+                target_root=self.get_agent_active_root(config_name, agent_id),
+                skill_sources=active_sources,
+            )
+
+            disabled_names = _dedupe_preserve_order(
+                self._resolve_disabled_skill_names(
+                    config_name=config_name,
+                    agent_id=agent_id,
+                    default_skills=agent_defaults.get(agent_id, []),
+                ),
+            )
+            disabled_sources = {
+                skill_name: self._resolve_agent_skill_source_dir(
+                    config_name=config_name,
+                    agent_id=agent_id,
+                    skill_name=skill_name,
+                )
+                for skill_name in disabled_names
+            }
+            self.sync_skill_dirs(
+                target_root=self.get_agent_disabled_root(config_name, agent_id),
+                skill_sources=disabled_sources,
+            )
+
+        if auto_reload:
+            self.watch_active_skills(config_name, agent_defaults)
+
+        return active_map
+
+    def _is_shared_skill(self, skill_name: str) -> bool:
+        try:
+            self._resolve_source_dir(skill_name)
+        except FileNotFoundError:
+            return False
+        return True
+
+    def watch_active_skills(
+        self,
+        config_name: str,
+        agent_defaults: Dict[str, Iterable[str]],
+        callback: Optional[Any] = None,
+    ) -> "_SkillsWatcher":
+        """Start file system monitoring on active skill directories.
+
+        Args:
+            config_name: Run configuration name.
+            agent_defaults: Map of agent_id -> default skill names.
+            callback: Optional callable invoked on file changes with
+                (changed_paths: List[Path]).
+
+        Returns:
+            A _SkillsWatcher instance. Call .stop() to halt monitoring.
+        """
+        if not WATCHDOG_AVAILABLE:
+            raise ImportError(
+                "watchdog is required for watch_active_skills. "
+                "Install it with: pip install watchdog"
+            )
+
+        watched_paths: List[Path] = []
+        for agent_id in agent_defaults:
+            active_root = self.get_agent_active_root(config_name, agent_id)
+            if active_root.exists():
+                watched_paths.append(active_root)
+            local_root = self.get_agent_local_root(config_name, agent_id)
+            if local_root.exists():
+                watched_paths.append(local_root)
+
+        handler = _SkillsChangeHandler(watched_paths, callback, self._lock)
+        observer = Observer()
+        for path in watched_paths:
+            observer.schedule(handler, str(path), recursive=True)
+        observer.start()
+        return _SkillsWatcher(observer, handler)
+
+    def reload_skills_if_changed(
+        self,
+        config_name: str,
+        agent_defaults: Dict[str, Iterable[str]],
+    ) -> Dict[str, List[Path]]:
+        """Check for file changes and reload active skills if needed.
+
+        Args:
+            config_name: Run configuration name.
+            agent_defaults: Map of agent_id -> default skill names.
+
+        Returns:
+            Map of agent_id -> list of reloaded skill paths, or empty dict
+            if no changes were detected.
+        """
+        with self._lock:
+            changed = self._pending_skill_changes.get(config_name)
+            if not changed:
+                return {}
+
+            self._pending_skill_changes[config_name] = set()
+
+        return self.prepare_active_skills(config_name, agent_defaults)
+
+    # -------------------------------------------------------------------------
+    # Internal change-tracking state (populated by _SkillsChangeHandler)
+    # -------------------------------------------------------------------------
+    _pending_skill_changes: Dict[str, Set[Path]] = {}
+
+    def _resolve_disabled_skill_names(
+        self,
+        config_name: str,
+        agent_id: str,
+        default_skills: Iterable[str],
+    ) -> List[str]:
+        """Resolve explicit disabled skills for one agent."""
+        bootstrap = get_bootstrap_config_for_run(self.project_root, config_name)
+        override = bootstrap.agent_override(agent_id)
+        baseline = list(override.get("skills", list(default_skills)))
+        agent_config = load_agent_workspace_config(
+            self.get_agent_asset_dir(config_name, agent_id) / "agent.yaml",
+        )
+        manifest = self.load_activation_manifest(config_name)
+        disabled = list(manifest.get("global_disabled_skills", []))
+        disabled.extend(manifest.get("agent_disabled_skills", {}).get(agent_id, []))
+        disabled.extend(agent_config.disabled_skills)
+        for skill_name in baseline:
+            if skill_name in agent_config.disabled_skills and skill_name not in disabled:
+                disabled.append(skill_name)
+        for item in self.list_agent_local_skills(config_name, agent_id):
+            if item.skill_name in agent_config.disabled_skills and item.skill_name not in disabled:
+                disabled.append(item.skill_name)
+        return [
+            skill
+            for skill in disabled
+            if self._skill_exists_for_agent(config_name, agent_id, skill)
+        ]
+
+
+class _SkillsWatcher:
+    """Handle returned by watch_active_skills; call .stop() to halt monitoring."""
+
+    def __init__(self, observer: Observer, handler: "_SkillsChangeHandler") -> None:
+        self._observer = observer
+        self._handler = handler
+
+    def stop(self) -> None:
+        """Stop the file system observer."""
+        self._observer.stop()
+        self._observer.join()
+
+
+class _SkillsChangeHandler(FileSystemEventHandler):
+    """Collects file-change events on skill directories."""
+
+    def __init__(
+        self,
+        watched_paths: List[Path],
+        callback: Optional[Any] = None,
+        lock: Optional[Lock] = None,
+    ) -> None:
+        super().__init__()
+        self._watched_paths = watched_paths
+        self._callback = callback
+        self._lock = lock
+
+    def on_any_event(self, event: FileSystemEvent) -> None:
+        if event.is_directory:
+            return
+        src_path = Path(event.src_path)
+        for watched in self._watched_paths:
+            if src_path.is_relative_to(watched):
+                run_id = self._run_id_from_path(src_path)
+                if self._lock:
+                    with self._lock:
+                        SkillsManager._pending_skill_changes.setdefault(
+                            run_id, set()
+                        ).add(src_path)
+                else:
+                    SkillsManager._pending_skill_changes.setdefault(
+                        run_id, set()
+                    ).add(src_path)
+                if self._callback:
+                    self._callback([src_path])
+                break
+
+    @staticmethod
+    def _run_id_from_path(path: Path) -> str:
+        """Infer config_name from a path like runs/{config_name}/skills/active/..."""
+        parts = path.parts
+        for i, part in enumerate(parts):
+            if part == "runs" and i + 1 < len(parts):
+                return parts[i + 1]
+        return "default"
+
+def _dedupe_preserve_order(items: Iterable[str]) -> List[str]:
+    result: List[str] = []
+    for item in items:
+        if item not in result:
+            result.append(item)
+    return result
+
+
+def _normalize_skill_name(raw_name: str) -> str:
+    normalized = str(raw_name or "").strip().lower().replace(" ", "_").replace("-", "_")
+    allowed = [ch for ch in normalized if ch.isalnum() or ch == "_"]
+    return "".join(allowed).strip("_")
--- a/backend/agents/team/init.py
+++ b/backend/agents/team/init.py
@@ -0,0 +1,18 @@
+# -*- coding: utf-8 -*-
+"""Team module for multi-agent orchestration.
+
+Provides inter-agent communication, task delegation, and coordination
+for subagent spawning and lifecycle management.
+"""
+
+from .messenger import AgentMessenger
+from .task_delegator import TaskDelegator
+from .team_coordinator import TeamCoordinator
+from .registry import AgentRegistry
+
+__all__ = [
+    "AgentMessenger",
+    "TaskDelegator",
+    "TeamCoordinator",
+    "AgentRegistry",
+]
--- a/backend/agents/team/messenger.py
+++ b/backend/agents/team/messenger.py
@@ -0,0 +1,225 @@
+# -*- coding: utf-8 -*-
+"""AgentMessenger - Pub/sub inter-agent communication.
+
+Provides broadcast(), send(), and subscribe() for message passing
+between agents using AgentScope's Msg format.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import logging
+from typing import Any, Callable, Dict, List, Optional, Set
+
+from agentscope.message import Msg
+
+logger = logging.getLogger(__name__)
+
+
+class AgentMessenger:
+    """Pub/sub messenger for inter-agent communication.
+
+    Supports:
+    - broadcast(): Send message to all subscribers
+    - send(): Send message to specific agent
+    - subscribe(): Register callback for agent messages
+    - announce(): Send system-wide announcement
+    - enable_auto_broadcast: Auto-broadcast agent replies to all participants
+
+    Messages use AgentScope's Msg format for compatibility.
+    """
+
+    def __init__(self, enable_auto_broadcast: bool = False):
+        """Initialize the messenger.
+
+        Args:
+            enable_auto_broadcast: If True, agent replies are automatically
+                broadcast to all subscribed agents.
+        """
+        self._subscriptions: Dict[str, List[Callable[[Msg], None]]] = {}
+        self._inbox: Dict[str, List[Msg]] = {}
+        self._locks: Dict[str, asyncio.Lock] = {}
+        self._enable_auto_broadcast = enable_auto_broadcast
+        self._participants: Set[str] = set()
+
+    def subscribe(
+        self,
+        agent_id: str,
+        callback: Callable[[Msg], None],
+    ) -> None:
+        """Subscribe an agent to receive messages.
+
+        Args:
+            agent_id: Target agent identifier
+            callback: Async function to call when message received
+        """
+        if agent_id not in self._subscriptions:
+            self._subscriptions[agent_id] = []
+        self._subscriptions[agent_id].append(callback)
+        logger.debug("Agent %s subscribed to messages", agent_id)
+
+    def unsubscribe(self, agent_id: str, callback: Callable[[Msg], None]) -> None:
+        """Unsubscribe an agent from messages.
+
+        Args:
+            agent_id: Target agent identifier
+            callback: Callback to remove
+        """
+        if agent_id in self._subscriptions:
+            try:
+                self._subscriptions[agent_id].remove(callback)
+                logger.debug("Agent %s unsubscribed from messages", agent_id)
+            except ValueError:
+                pass
+
+    async def send(
+        self,
+        to_agent: str,
+        message: Msg,
+    ) -> None:
+        """Send message to specific agent.
+
+        Args:
+            to_agent: Target agent identifier
+            message: Message to send (uses Msg format)
+        """
+        async def _deliver():
+            if to_agent in self._subscriptions:
+                for callback in self._subscriptions[to_agent]:
+                    try:
+                        if asyncio.iscoroutinefunction(callback):
+                            await callback(message)
+                        else:
+                            callback(message)
+                    except Exception as e:
+                        logger.error(
+                            "Error delivering message to %s: %s",
+                            to_agent,
+                            e,
+                        )
+
+        await _deliver()
+
+    async def broadcast(self, message: Msg) -> None:
+        """Broadcast message to all subscribed agents.
+
+        Args:
+            message: Message to broadcast (uses Msg format)
+        """
+        delivery_tasks = []
+        for agent_id, callbacks in self._subscriptions.items():
+            for callback in callbacks:
+                async def _deliver(cb=callback, aid=agent_id):
+                    try:
+                        if asyncio.iscoroutinefunction(cb):
+                            await cb(message)
+                        else:
+                            cb(message)
+                    except Exception as e:
+                        logger.error(
+                            "Error broadcasting to %s: %s",
+                            aid,
+                            e,
+                        )
+                delivery_tasks.append(_deliver())
+
+        if delivery_tasks:
+            await asyncio.gather(*delivery_tasks)
+
+    def inbox(self, agent_id: str) -> List[Msg]:
+        """Get and clear inbox for agent.
+
+        Args:
+            agent_id: Agent identifier
+
+        Returns:
+            List of messages in inbox
+        """
+        messages = self._inbox.get(agent_id, [])
+        self._inbox[agent_id] = []
+        return messages
+
+    def inbox_count(self, agent_id: str) -> int:
+        """Count messages in agent's inbox without clearing.
+
+        Args:
+            agent_id: Agent identifier
+
+        Returns:
+            Number of messages waiting
+        """
+        return len(self._inbox.get(agent_id, []))
+
+    def add_participant(self, agent_id: str) -> None:
+        """Add a participant to the messenger.
+
+        Participants are the agents that can receive auto-broadcast messages.
+
+        Args:
+            agent_id: Agent identifier to add
+        """
+        self._participants.add(agent_id)
+        logger.debug("Agent %s added as participant", agent_id)
+
+    def remove_participant(self, agent_id: str) -> None:
+        """Remove a participant from the messenger.
+
+        Args:
+            agent_id: Agent identifier to remove
+        """
+        self._participants.discard(agent_id)
+        logger.debug("Agent %s removed from participants", agent_id)
+
+    @property
+    def enable_auto_broadcast(self) -> bool:
+        """Check if auto_broadcast is enabled."""
+        return self._enable_auto_broadcast
+
+    @enable_auto_broadcast.setter
+    def enable_auto_broadcast(self, value: bool) -> None:
+        """Enable or disable auto_broadcast."""
+        self._enable_auto_broadcast = value
+        logger.debug("Auto_broadcast set to %s", value)
+
+    async def announce(self, message: Msg) -> None:
+        """Send a system-wide announcement to all participants.
+
+        Unlike broadcast(), announce() sends a message from the system/host
+        to all participants without requiring prior subscription.
+
+        Args:
+            message: Announcement message (uses Msg format)
+        """
+        logger.info("System announcement: %s", message.content)
+        await self.broadcast(message)
+
+    async def auto_broadcast(self, message: Msg) -> None:
+        """Auto-broadcast message to all participants.
+
+        This is called internally when enable_auto_broadcast is True.
+        Broadcasts to all registered participants.
+
+        Args:
+            message: Message to auto-broadcast (uses Msg format)
+        """
+        if not self._enable_auto_broadcast:
+            return
+
+        # Broadcast to all participants
+        for participant_id in self._participants:
+            if participant_id in self._subscriptions:
+                for callback in self._subscriptions[participant_id]:
+                    try:
+                        if asyncio.iscoroutinefunction(callback):
+                            await callback(message)
+                        else:
+                            callback(message)
+                    except Exception as e:
+                        logger.error(
+                            "Error auto-broadcasting to %s: %s",
+                            participant_id,
+                            e,
+                        )
+
+
+__all__ = ["AgentMessenger"]
--- a/backend/agents/team/registry.py
+++ b/backend/agents/team/registry.py
@@ -0,0 +1,188 @@
+# -*- coding: utf-8 -*-
+"""AgentRegistry - Agent registration and lookup by role.
+
+Provides register(), unregister(), and get_by_role() for agent
+discovery and management.
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import Any, Dict, List, Optional
+
+from agentscope.message import Msg
+
+logger = logging.getLogger(__name__)
+
+
+class AgentRegistry:
+    """Registry for agent instances with role-based lookup.
+
+    Supports:
+    - register(): Add agent with roles
+    - unregister(): Remove agent
+    - get_by_role(): Find agents by role
+    - get_by_id(): Get specific agent
+
+    Each agent can have multiple roles for flexible dispatch.
+    """
+
+    def __init__(self):
+        self._agents: Dict[str, Any] = {}
+        self._roles: Dict[str, List[str]] = {}
+        self._agent_roles: Dict[str, List[str]] = {}
+
+    def register(
+        self,
+        agent_id: str,
+        agent: Any,
+        roles: Optional[List[str]] = None,
+    ) -> None:
+        """Register an agent with optional roles.
+
+        Args:
+            agent_id: Unique agent identifier
+            agent: Agent instance
+            roles: Optional list of role strings
+        """
+        self._agents[agent_id] = agent
+        self._agent_roles[agent_id] = roles or []
+
+        for role in self._agent_roles[agent_id]:
+            if role not in self._roles:
+                self._roles[role] = []
+            if agent_id not in self._roles[role]:
+                self._roles[role].append(agent_id)
+
+        logger.info(
+            "Registered agent %s with roles %s",
+            agent_id,
+            self._agent_roles[agent_id],
+        )
+
+    def unregister(self, agent_id: str) -> bool:
+        """Unregister an agent.
+
+        Args:
+            agent_id: Agent identifier to remove
+
+        Returns:
+            True if agent was removed
+        """
+        if agent_id not in self._agents:
+            return False
+
+        roles = self._agent_roles.pop(agent_id, [])
+        for role in roles:
+            if role in self._roles:
+                try:
+                    self._roles[role].remove(agent_id)
+                except ValueError:
+                    pass
+
+        del self._agents[agent_id]
+        logger.info("Unregistered agent: %s", agent_id)
+        return True
+
+    def get_by_id(self, agent_id: str) -> Optional[Any]:
+        """Get agent by ID.
+
+        Args:
+            agent_id: Agent identifier
+
+        Returns:
+            Agent instance or None
+        """
+        return self._agents.get(agent_id)
+
+    def get_by_role(self, role: str) -> List[Any]:
+        """Get all agents with a given role.
+
+        Args:
+            role: Role string to search for
+
+        Returns:
+            List of agent instances with the role
+        """
+        agent_ids = self._roles.get(role, [])
+        return [self._agents[aid] for aid in agent_ids if aid in self._agents]
+
+    def get_by_roles(self, roles: List[str]) -> List[Any]:
+        """Get agents matching ANY of the given roles.
+
+        Args:
+            roles: List of role strings
+
+        Returns:
+            List of unique agent instances matching any role
+        """
+        seen = set()
+        result = []
+        for role in roles:
+            for agent in self.get_by_role(role):
+                if id(agent) not in seen:
+                    seen.add(id(agent))
+                    result.append(agent)
+        return result
+
+    def list_agents(self) -> List[str]:
+        """List all registered agent IDs.
+
+        Returns:
+            List of agent identifiers
+        """
+        return list(self._agents.keys())
+
+    def list_roles(self) -> List[str]:
+        """List all registered roles.
+
+        Returns:
+            List of role strings
+        """
+        return list(self._roles.keys())
+
+    def list_roles_for_agent(self, agent_id: str) -> List[str]:
+        """List roles for specific agent.
+
+        Args:
+            agent_id: Agent identifier
+
+        Returns:
+            List of role strings
+        """
+        return list(self._agent_roles.get(agent_id, []))
+
+    def update_roles(self, agent_id: str, roles: List[str]) -> None:
+        """Update roles for an existing agent.
+
+        Args:
+            agent_id: Agent identifier
+            roles: New list of roles
+        """
+        if agent_id not in self._agents:
+            raise KeyError(f"Agent not registered: {agent_id}")
+
+        old_roles = self._agent_roles.get(agent_id, [])
+        for role in old_roles:
+            if role in self._roles:
+                try:
+                    self._roles[role].remove(agent_id)
+                except ValueError:
+                    pass
+
+        self._agent_roles[agent_id] = roles
+        for role in roles:
+            if role not in self._roles:
+                self._roles[role] = []
+            if agent_id not in self._roles[role]:
+                self._roles[role].append(agent_id)
+
+        logger.info("Updated roles for agent %s: %s", agent_id, roles)
+
+    @property
+    def agents(self) -> Dict[str, Any]:
+        """Get copy of registered agents dict."""
+        return dict(self._agents)
+
+
+__all__ = ["AgentRegistry"]
--- a/backend/agents/team/task_delegator.py
+++ b/backend/agents/team/task_delegator.py
@@ -0,0 +1,620 @@
+# -*- coding: utf-8 -*-
+"""TaskDelegator - Subagent spawning and task delegation.
+
+Provides delegate() and delegate_parallel() for spawning subagents
+with separate context and memory. Supports runtime dynamic subagent
+definition via task_data with description, prompt, and tools.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import logging
+import uuid
+from typing import Any, Awaitable, Callable, Dict, List, Optional, Union
+
+from agentscope.message import Msg
+
+logger = logging.getLogger(__name__)
+
+# Default timeout for subagent execution (seconds)
+DEFAULT_EXECUTION_TIMEOUT = 120.0
+
+
+# Type alias for subagent specification
+SubagentSpec = Dict[str, Any]
+"""Subagent specification format:
+{
+    "description": "Expert code reviewer...",
+    "prompt": "Analyze code quality...",
+    "tools": ["Read", "Glob", "Grep"],  # Optional: list of tool names
+    "model": "gpt-4o",  # Optional: model name
+}
+"""
+
+
+class TaskDelegator:
+    """Delegates tasks to subagents with isolated context.
+
+    Supports:
+    - delegate(): Spawn single subagent for task
+    - delegate_parallel(): Spawn multiple subagents concurrently
+    - delegate_task(): Delegate with dynamic subagent definition from task_data
+
+    Each subagent gets its own memory/context to prevent
+    cross-contamination.
+
+    Dynamic Subagent Definition:
+        task_data can include an "agents" dict to define subagents inline:
+
+        task_data = {
+            "task": "Review the code changes",
+            "agents": {
+                "code-reviewer": {
+                    "description": "Expert code reviewer for quality and security.",
+                    "prompt": "Analyze code quality and suggest improvements.",
+                    "tools": ["Read", "Glob", "Grep"],
+                }
+            }
+        }
+    """
+
+    def __init__(self, agent: Any):
+        """Initialize TaskDelegator.
+
+        Args:
+            agent: Parent EvoAgent instance for accessing model, formatter, workspace
+        """
+        self._agent = agent
+        # Get messenger from parent agent if available
+        self._messenger = getattr(agent, "messenger", None)
+        self._registry = getattr(agent, "_registry", None)
+        self._subagents: Dict[str, Any] = {}
+        self._dynamic_subagents: Dict[str, SubagentSpec] = {}
+        self._tasks: Dict[str, asyncio.Task] = {}
+
+        # Extract model and formatter from parent agent
+        self._model = getattr(agent, "model", None)
+        self._formatter = getattr(agent, "formatter", None)
+        self._workspace_dir = getattr(agent, "workspace_dir", None)
+        self._config_name = getattr(agent, "config_name", None)
+
+    async def delegate(
+        self,
+        agent_id: str,
+        task: Callable[..., Awaitable[Msg]],
+        context: Optional[Dict[str, Any]] = None,
+    ) -> asyncio.Task:
+        """Delegate task to a single subagent.
+
+        Args:
+            agent_id: Unique identifier for this subagent instance
+            task: Async function representing the task
+            context: Optional context dict for the subagent
+
+        Returns:
+            asyncio.Task for the delegated task
+        """
+        async def _run_with_context():
+            result = await task(context or {})
+            return result
+
+        self._tasks[agent_id] = asyncio.create_task(_run_with_context())
+        logger.info("Delegated task to subagent: %s", agent_id)
+        return self._tasks[agent_id]
+
+    async def delegate_parallel(
+        self,
+        tasks: List[Dict[str, Any]],
+    ) -> List[asyncio.Task]:
+        """Delegate multiple tasks in parallel.
+
+        Args:
+            tasks: List of task dicts with keys:
+                - agent_id: Unique identifier
+                - task: Async function to execute
+                - context: Optional context dict
+
+        Returns:
+            List of asyncio.Task for all delegated tasks
+        """
+        async def _run_task(task_def: Dict[str, Any]):
+            agent_id = task_def["agent_id"]
+            task_func = task_def["task"]
+            context = task_def.get("context", {})
+
+            async def _run_with_context():
+                return await task_func(context)
+
+            self._tasks[agent_id] = asyncio.create_task(_run_with_context())
+            return self._tasks[agent_id]
+
+        gathered_tasks = await asyncio.gather(
+            *[_run_task(t) for t in tasks],
+            return_exceptions=True,
+        )
+
+        valid_tasks = [t for t in gathered_tasks if isinstance(t, asyncio.Task)]
+        logger.info(
+            "Delegated %d tasks in parallel (%d succeeded)",
+            len(tasks),
+            len(valid_tasks),
+        )
+        return valid_tasks
+
+    async def wait_for(self, agent_id: str, timeout: Optional[float] = None) -> Any:
+        """Wait for subagent task to complete.
+
+        Args:
+            agent_id: Subagent identifier
+            timeout: Optional timeout in seconds
+
+        Returns:
+            Task result
+
+        Raises:
+            asyncio.TimeoutError: If task doesn't complete in time
+            KeyError: If agent_id not found
+        """
+        if agent_id not in self._tasks:
+            raise KeyError(f"Unknown subagent: {agent_id}")
+
+        try:
+            return await asyncio.wait_for(
+                self._tasks[agent_id],
+                timeout=timeout,
+            )
+        except asyncio.TimeoutError:
+            logger.warning("Task %s timed out after %s seconds", agent_id, timeout)
+            raise
+
+    async def cancel(self, agent_id: str) -> bool:
+        """Cancel a subagent task.
+
+        Args:
+            agent_id: Subagent identifier
+
+        Returns:
+            True if task was cancelled
+        """
+        if agent_id in self._tasks:
+            self._tasks[agent_id].cancel()
+            del self._tasks[agent_id]
+            logger.info("Cancelled subagent task: %s", agent_id)
+            return True
+        return False
+
+    def list_tasks(self) -> List[str]:
+        """List active subagent task IDs.
+
+        Returns:
+            List of agent_ids with pending tasks
+        """
+        return list(self._tasks.keys())
+
+    @property
+    def tasks(self) -> Dict[str, asyncio.Task]:
+        """Get copy of active tasks dict."""
+        return dict(self._tasks)
+
+    async def delegate_task(
+        self,
+        task_type: str,
+        task_data: Dict[str, Any],
+        target_agent: Optional[str] = None,
+    ) -> Dict[str, Any]:
+        """Delegate a task with optional dynamic subagent definition.
+
+        Supports runtime subagent definition via task_data["agents"]:
+
+            task_data = {
+                "task": "Review code changes",
+                "agents": {
+                    "code-reviewer": {
+                        "description": "Expert code reviewer...",
+                        "prompt": "Analyze code quality...",
+                        "tools": ["Read", "Glob", "Grep"],
+                    }
+                }
+            }
+
+        Args:
+            task_type: Type of task (e.g., "analysis", "review", "research")
+            task_data: Task payload, may include "agents" for dynamic subagent def
+            target_agent: Optional specific agent ID to delegate to
+
+        Returns:
+            Dict with "success" and result/error
+        """
+        try:
+            # Extract dynamic subagent definitions from task_data
+            agents_def = task_data.get("agents", {})
+
+            if agents_def:
+                # Register dynamic subagents
+                for agent_name, agent_spec in agents_def.items():
+                    self._dynamic_subagents[agent_name] = agent_spec
+                    logger.info(
+                        "Registered dynamic subagent: %s (description: %s)",
+                        agent_name,
+                        agent_spec.get("description", "")[:50],
+                    )
+
+            # Determine target agent
+            effective_target = target_agent
+            if not effective_target:
+                # Use first available dynamic subagent or default
+                if agents_def:
+                    effective_target = next(iter(agents_def.keys()))
+                else:
+                    effective_target = "default"
+
+            # Execute the task (async)
+            task_result = await self._execute_task(
+                task_type=task_type,
+                task_data=task_data,
+                target_agent=effective_target,
+            )
+
+            # Clean up dynamic subagents after execution
+            for agent_name in agents_def.keys():
+                self._dynamic_subagents.pop(agent_name, None)
+
+            return {
+                "success": True,
+                "result": task_result,
+                "subagents_used": list(agents_def.keys()) if agents_def else [],
+            }
+
+        except Exception as e:
+            logger.error("Task delegation failed: %s", e)
+            return {
+                "success": False,
+                "error": str(e),
+            }
+
+    async def _execute_task(
+        self,
+        task_type: str,
+        task_data: Dict[str, Any],
+        target_agent: str,
+    ) -> Dict[str, Any]:
+        """Execute the delegated task with a real subagent.
+
+        Args:
+            task_type: Type of task
+            task_data: Task payload
+            target_agent: Target agent identifier
+
+        Returns:
+            Task execution result with success/failure info
+        """
+        task_content = task_data.get("task", task_data.get("prompt", ""))
+        timeout = task_data.get("timeout", DEFAULT_EXECUTION_TIMEOUT)
+
+        # Check if we have a dynamic subagent spec for this target
+        agent_spec = self._dynamic_subagents.get(target_agent)
+
+        if agent_spec:
+            logger.info(
+                "Executing task '%s' with dynamic subagent '%s'",
+                task_type,
+                target_agent,
+            )
+            return await self._create_and_run_subagent(
+                agent_name=target_agent,
+                agent_spec=agent_spec,
+                task_content=task_content,
+                task_type=task_type,
+                timeout=timeout,
+            )
+
+        # Fallback: try to use parent agent's model to process the task directly
+        logger.info(
+            "Executing task '%s' with parent agent '%s' (no dynamic subagent)",
+            task_type,
+            target_agent,
+        )
+        return await self._run_with_parent_agent(
+            task_content=task_content,
+            task_type=task_type,
+            timeout=timeout,
+        )
+
+    async def _create_and_run_subagent(
+        self,
+        agent_name: str,
+        agent_spec: SubagentSpec,
+        task_content: str,
+        task_type: str,
+        timeout: float,
+    ) -> Dict[str, Any]:
+        """Create and run a dynamic subagent.
+
+        Args:
+            agent_name: Name identifier for the subagent
+            agent_spec: Subagent specification (description, prompt, tools, model)
+            task_content: Task prompt to send to the subagent
+            task_type: Type of task
+            timeout: Execution timeout in seconds
+
+        Returns:
+            Dict with execution results
+        """
+        subagent_id = f"subagent_{agent_name}_{uuid.uuid4().hex[:8]}"
+
+        try:
+            # Create subagent instance
+            subagent = await self._create_subagent(
+                subagent_id=subagent_id,
+                agent_spec=agent_spec,
+            )
+
+            if subagent is None:
+                return {
+                    "task_type": task_type,
+                    "task": task_content,
+                    "subagent": agent_name,
+                    "status": "failed",
+                    "error": "Failed to create subagent",
+                    "message": f"Could not instantiate subagent '{agent_name}'",
+                }
+
+            # Store for potential cleanup
+            self._subagents[subagent_id] = subagent
+
+            # Execute with timeout
+            result = await asyncio.wait_for(
+                self._run_subagent(subagent, task_content),
+                timeout=timeout,
+            )
+
+            # Extract response content
+            response_content = ""
+            if isinstance(result, Msg):
+                response_content = result.content
+            elif hasattr(result, "content"):
+                response_content = str(result.content)
+            elif isinstance(result, dict):
+                response_content = result.get("content", str(result))
+            else:
+                response_content = str(result)
+
+            logger.info(
+                "Subagent '%s' completed task '%s' successfully",
+                agent_name,
+                task_type,
+            )
+
+            return {
+                "task_type": task_type,
+                "task": task_content,
+                "subagent": {
+                    "name": agent_name,
+                    "id": subagent_id,
+                    "description": agent_spec.get("description", ""),
+                },
+                "status": "completed",
+                "response": response_content,
+                "message": f"Task '{task_type}' executed with subagent '{agent_name}'",
+            }
+
+        except asyncio.TimeoutError:
+            logger.warning(
+                "Subagent '%s' timed out after %.1f seconds for task '%s'",
+                agent_name,
+                timeout,
+                task_type,
+            )
+            # Cancel the task if still running
+            if subagent_id in self._subagents:
+                self._subagents.pop(subagent_id, None)
+            return {
+                "task_type": task_type,
+                "task": task_content,
+                "subagent": agent_name,
+                "status": "timeout",
+                "error": f"Execution timed out after {timeout} seconds",
+                "message": f"Task '{task_type}' timed out for subagent '{agent_name}'",
+            }
+
+        except Exception as e:
+            logger.error(
+                "Subagent '%s' failed for task '%s': %s",
+                agent_name,
+                task_type,
+                e,
+                exc_info=True,
+            )
+            # Cleanup on failure
+            if subagent_id in self._subagents:
+                self._subagents.pop(subagent_id, None)
+            return {
+                "task_type": task_type,
+                "task": task_content,
+                "subagent": agent_name,
+                "status": "error",
+                "error": str(e),
+                "message": f"Task '{task_type}' failed for subagent '{agent_name}': {e}",
+            }
+
+    async def _create_subagent(
+        self,
+        subagent_id: str,
+        agent_spec: SubagentSpec,
+    ) -> Optional[Any]:
+        """Create a subagent instance.
+
+        Uses the parent agent's model/formatter to create a lightweight
+        subagent for task execution.
+
+        Args:
+            subagent_id: Unique identifier for the subagent
+            agent_spec: Subagent specification
+
+        Returns:
+            Subagent instance or None if creation fails
+        """
+        try:
+            # Import here to avoid circular imports
+            from agentscope.memory import InMemoryMemory
+
+            # Get model and formatter from parent
+            model = self._model
+            formatter = self._formatter
+
+            if model is None:
+                logger.error("Cannot create subagent: parent agent has no model")
+                return None
+
+            # Build system prompt from agent spec
+            description = agent_spec.get("description", "")
+            prompt_template = agent_spec.get("prompt", "")
+            system_prompt = f"""You are {description}
+
+{prompt_template}
+
+Your task is to complete the user's request below.
+"""
+
+            # Create a minimal ReActAgent as the subagent
+            from agentscope.agent import ReActAgent
+
+            subagent = ReActAgent(
+                name=subagent_id,
+                model=model,
+                sys_prompt=system_prompt,
+                toolkit=None,  # Could load tools from agent_spec.get("tools", [])
+                memory=InMemoryMemory(),
+                formatter=formatter,
+                max_iters=agent_spec.get("max_iters", 5),
+            )
+
+            logger.debug("Created subagent: %s", subagent_id)
+            return subagent
+
+        except Exception as e:
+            logger.error(
+                "Failed to create subagent '%s': %s",
+                subagent_id,
+                e,
+                exc_info=True,
+            )
+            return None
+
+    async def _run_subagent(
+        self,
+        subagent: Any,
+        task_content: str,
+    ) -> Any:
+        """Run a subagent with the given task.
+
+        Args:
+            subagent: Subagent instance
+            task_content: Task prompt
+
+        Returns:
+            Agent response (Msg or similar)
+        """
+        from agentscope.message import Msg
+
+        # Create message for the subagent
+        task_msg = Msg(
+            name="user",
+            content=task_content,
+            role="user",
+        )
+
+        # Execute the agent
+        response = await subagent.reply(task_msg)
+        return response
+
+    async def _run_with_parent_agent(
+        self,
+        task_content: str,
+        task_type: str,
+        timeout: float,
+    ) -> Dict[str, Any]:
+        """Run task using the parent agent directly.
+
+        Used when no dynamic subagent is defined.
+
+        Args:
+            task_content: Task prompt
+            task_type: Type of task
+            timeout: Execution timeout
+
+        Returns:
+            Dict with execution results
+        """
+        try:
+            result = await asyncio.wait_for(
+                self._agent.reply(Msg(
+                    name="user",
+                    content=task_content,
+                    role="user",
+                )),
+                timeout=timeout,
+            )
+
+            response_content = ""
+            if isinstance(result, Msg):
+                response_content = result.content
+            elif hasattr(result, "content"):
+                response_content = str(result.content)
+            else:
+                response_content = str(result)
+
+            return {
+                "task_type": task_type,
+                "task": task_content,
+                "status": "completed",
+                "response": response_content,
+                "message": f"Task '{task_type}' executed with parent agent",
+            }
+
+        except asyncio.TimeoutError:
+            return {
+                "task_type": task_type,
+                "task": task_content,
+                "status": "timeout",
+                "error": f"Execution timed out after {timeout} seconds",
+                "message": f"Task '{task_type}' timed out",
+            }
+
+        except Exception as e:
+            logger.error(
+                "Parent agent failed for task '%s': %s",
+                task_type,
+                e,
+                exc_info=True,
+            )
+            return {
+                "task_type": task_type,
+                "task": task_content,
+                "status": "error",
+                "error": str(e),
+                "message": f"Task '{task_type}' failed: {e}",
+            }
+
+    def get_dynamic_subagent(self, name: str) -> Optional[SubagentSpec]:
+        """Get a dynamically defined subagent specification.
+
+        Args:
+            name: Subagent name
+
+        Returns:
+            Subagent spec dict or None if not found
+        """
+        return self._dynamic_subagents.get(name)
+
+    def list_dynamic_subagents(self) -> List[str]:
+        """List all registered dynamic subagent names.
+
+        Returns:
+            List of subagent names
+        """
+        return list(self._dynamic_subagents.keys())
+
+
+__all__ = ["TaskDelegator", "SubagentSpec"]
--- a/backend/agents/team/team_coordinator.py
+++ b/backend/agents/team/team_coordinator.py
@@ -0,0 +1,389 @@
+# -*- coding: utf-8 -*-
+"""TeamCoordinator - Agent lifecycle management and execution.
+
+Provides run_parallel() using asyncio.gather() and run_sequential()
+for coordinating multiple agents.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import logging
+from typing import Any, Awaitable, Callable, Dict, List, Optional, Type
+
+from agentscope.message import Msg
+
+logger = logging.getLogger(__name__)
+
+
+class TeamCoordinator:
+    """Coordinates agent lifecycle and execution.
+
+    Supports:
+    - run_parallel(): Execute multiple agents concurrently with asyncio.gather()
+    - run_sequential(): Execute agents one after another
+    - run_phase(): Execute a named phase with registered agents
+    - register_agent(): Add agent to coordinator
+    - unregister_agent(): Remove agent from coordinator
+
+    Each agent maintains separate context/memory.
+    """
+
+    def __init__(
+        self,
+        participants: Optional[List[Any]] = None,
+        task_content: Optional[str] = None,
+        messenger: Optional[Any] = None,
+        registry: Optional[Any] = None,
+    ):
+        """Initialize TeamCoordinator.
+
+        Args:
+            participants: List of agent instances to coordinate
+            task_content: Task description content for the agents
+            messenger: AgentMessenger for communication (optional)
+            registry: AgentRegistry for agent lookup (optional)
+        """
+        self._participants = participants or []
+        self._task_content = task_content or ""
+        self._messenger = messenger
+        self._registry = registry
+        self._agents: Dict[str, Any] = {}
+        self._running_tasks: Dict[str, asyncio.Task] = {}
+        # Auto-register participants
+        for agent in self._participants:
+            if hasattr(agent, "name"):
+                self._agents[agent.name] = agent
+            elif hasattr(agent, "id"):
+                self._agents[agent.id] = agent
+
+    def register_agent(self, agent_id: str, agent: Any) -> None:
+        """Register an agent with the coordinator.
+
+        Args:
+            agent_id: Unique agent identifier
+            agent: Agent instance
+        """
+        self._agents[agent_id] = agent
+        logger.info("Registered agent: %s", agent_id)
+
+    def unregister_agent(self, agent_id: str) -> None:
+        """Unregister an agent from the coordinator.
+
+        Args:
+            agent_id: Agent identifier to remove
+        """
+        if agent_id in self._agents:
+            del self._agents[agent_id]
+            logger.info("Unregistered agent: %s", agent_id)
+
+    def get_agent(self, agent_id: str) -> Any:
+        """Get registered agent by ID.
+
+        Args:
+            agent_id: Agent identifier
+
+        Returns:
+            Agent instance
+        """
+        return self._agents.get(agent_id)
+
+    def list_agents(self) -> List[str]:
+        """List all registered agent IDs.
+
+        Returns:
+            List of agent identifiers
+        """
+        return list(self._agents.keys())
+
+    async def run_parallel(
+        self,
+        agent_ids: List[str],
+        initial_message: Optional[Msg] = None,
+    ) -> Dict[str, Any]:
+        """Run multiple agents in parallel using asyncio.gather().
+
+        Args:
+            agent_ids: List of agent IDs to run concurrently
+            initial_message: Optional initial message to broadcast
+
+        Returns:
+            Dict mapping agent_id to result
+        """
+        async def _run_agent(aid: str) -> tuple[str, Any]:
+            agent = self._agents.get(aid)
+            if agent is None:
+                logger.error("Agent %s not found", aid)
+                return (aid, None)
+
+            try:
+                if hasattr(agent, "reply") and asyncio.iscoroutinefunction(agent.reply):
+                    if initial_message:
+                        result = await agent.reply(initial_message)
+                    else:
+                        result = await agent.reply()
+                elif hasattr(agent, "run") and asyncio.iscoroutinefunction(agent.run):
+                    result = await agent.run()
+                else:
+                    result = await agent()
+                logger.info("Agent %s completed successfully", aid)
+                return (aid, result)
+            except Exception as e:
+                logger.error("Agent %s failed: %s", aid, e)
+                return (aid, {"error": str(e)})
+
+        results = await asyncio.gather(
+            *[_run_agent(aid) for aid in agent_ids],
+            return_exceptions=True,
+        )
+
+        output: Dict[str, Any] = {}
+        for result in results:
+            if isinstance(result, tuple):
+                agent_id, agent_result = result
+                output[agent_id] = agent_result
+            else:
+                logger.error("Unexpected result from asyncio.gather: %s", result)
+
+        logger.info("Parallel run completed for %d agents", len(agent_ids))
+        return output
+
+    async def run_sequential(
+        self,
+        agent_ids: List[str],
+        initial_message: Optional[Msg] = None,
+    ) -> Dict[str, Any]:
+        """Run agents one after another in order.
+
+        Args:
+            agent_ids: List of agent IDs to run in sequence
+            initial_message: Optional initial message for first agent
+
+        Returns:
+            Dict mapping agent_id to result
+        """
+        output: Dict[str, Any] = {}
+        current_message = initial_message
+
+        for agent_id in agent_ids:
+            agent = self._agents.get(agent_id)
+            if agent is None:
+                logger.error("Agent %s not found", agent_id)
+                output[agent_id] = {"error": "Agent not found"}
+                continue
+
+            try:
+                if hasattr(agent, "reply") and asyncio.iscoroutinefunction(agent.reply):
+                    result = await agent.reply(current_message)
+                elif hasattr(agent, "run") and asyncio.iscoroutinefunction(agent.run):
+                    result = await agent.run()
+                else:
+                    result = await agent()
+
+                output[agent_id] = result
+                current_message = result
+                logger.info("Agent %s completed sequentially", agent_id)
+
+            except Exception as e:
+                logger.error("Agent %s failed: %s", agent_id, e)
+                output[agent_id] = {"error": str(e)}
+                break
+
+        logger.info("Sequential run completed for %d agents", len(agent_ids))
+        return output
+
+    async def run_phase(
+        self,
+        phase_name: str,
+        agent_ids: Optional[List[str]] = None,
+        metadata: Optional[Dict[str, Any]] = None,
+    ) -> List[Any]:
+        """Execute a named phase with registered agents.
+
+        Args:
+            phase_name: Name of the phase (e.g., "analyst_analysis")
+            agent_ids: Optional list of agent IDs; if None, uses all registered
+            metadata: Optional metadata to include in the message (e.g., tickers, date)
+
+        Returns:
+            List of results from each agent
+        """
+        if agent_ids is None:
+            agent_ids = list(self._agents.keys())
+
+        _agent_ids = [aid for aid in agent_ids if aid in self._agents]
+
+        logger.info(
+            "Running phase '%s' with %d agents: %s",
+            phase_name,
+            len(_agent_ids),
+            _agent_ids,
+        )
+
+        # Create messages for each agent
+        results: List[Any] = []
+        for agent_id in _agent_ids:
+            agent = self._agents[agent_id]
+            try:
+                if hasattr(agent, "reply") and asyncio.iscoroutinefunction(agent.reply):
+                    # Create a message for the agent with proper structure
+                    msg = Msg(
+                        name="system",
+                        content=self._task_content or f"Please execute phase: {phase_name}",
+                        role="user",
+                        metadata=metadata,
+                    )
+                    result = await agent.reply(msg)
+                elif hasattr(agent, "run") and asyncio.iscoroutinefunction(agent.run):
+                    result = await agent.run()
+                else:
+                    result = await agent()
+                results.append(result)
+                logger.info("Phase '%s': Agent %s completed", phase_name, agent_id)
+            except Exception as e:
+                logger.error("Phase '%s': Agent %s failed: %s", phase_name, agent_id, e)
+                results.append(None)
+
+        logger.info("Phase '%s' completed with %d results", phase_name, len(results))
+        return results
+
+    async def run_with_dependencies(
+        self,
+        agent_tasks: Dict[str, List[str]],
+        initial_message: Optional[Msg] = None,
+    ) -> Dict[str, Any]:
+        """Run agents respecting dependency graph.
+
+        Args:
+            agent_tasks: Dict mapping agent_id to list of prerequisite agent_ids
+            initial_message: Optional initial message
+
+        Returns:
+            Dict mapping agent_id to result
+        """
+        completed: Dict[str, Any] = {}
+        remaining = set(agent_tasks.keys())
+
+        while remaining:
+            ready = [
+                aid for aid in remaining
+                if all(dep in completed for dep in agent_tasks.get(aid, []))
+            ]
+
+            if not ready:
+                logger.error("Circular dependency detected in agent tasks")
+                for aid in remaining:
+                    completed[aid] = {"error": "Circular dependency"}
+                break
+
+            results = await self.run_parallel(ready, initial_message)
+            completed.update(results)
+
+            for aid in ready:
+                remaining.discard(aid)
+                initial_message = results.get(aid)
+
+        return completed
+
+    async def fanout_pipeline(
+        self,
+        agents: List[Any],
+        msg: Optional[Msg] = None,
+    ) -> List[Msg]:
+        """Fanout a message to multiple agents concurrently and collect all responses.
+
+        Similar to AgentScope's fanout_pipeline, this sends the same message
+        to all specified agents and returns a list of all agent responses.
+
+        Args:
+            agents: List of agent instances to fanout the message to
+            msg: Message to send to all agents (optional)
+
+        Returns:
+            List of Msg responses from each agent (in the same order as input agents)
+
+        Example:
+            >>> responses = await fanout_pipeline(
+            ...     agents=[alice, bob, charlie],
+            ...     msg=question,
+            ... )
+            >>> # responses is a list of Msg responses from each agent
+        """
+        async def _fanout_to_agent(agent: Any) -> Optional[Msg]:
+            """Send message to a single agent and return its response."""
+            try:
+                if hasattr(agent, "reply") and asyncio.iscoroutinefunction(agent.reply):
+                    result = await agent.reply(msg) if msg is not None else await agent.reply()
+                elif hasattr(agent, "run") and asyncio.iscoroutinefunction(agent.run):
+                    result = await agent.run()
+                else:
+                    result = await agent()
+
+                # Convert result to Msg if needed
+                if result is None:
+                    return None
+                if isinstance(result, Msg):
+                    return result
+                # If result is a dict with content, wrap it
+                if isinstance(result, dict) and "content" in result:
+                    return Msg(
+                        name=getattr(agent, "name", "unknown"),
+                        content=result.get("content", ""),
+                        role="assistant",
+                        metadata=result.get("metadata"),
+                    )
+                # Otherwise wrap the result
+                return Msg(
+                    name=getattr(agent, "name", "unknown"),
+                    content=str(result),
+                    role="assistant",
+                )
+            except Exception as e:
+                logger.error("Agent %s failed in fanout_pipeline: %s",
+                            getattr(agent, "name", "unknown"), e)
+                return None
+
+        # Run all agents concurrently
+        results = await asyncio.gather(
+            *[_fanout_to_agent(agent) for agent in agents],
+            return_exceptions=True,
+        )
+
+        # Filter out exceptions and keep only valid responses
+        responses: List[Msg] = []
+        for i, result in enumerate(results):
+            if isinstance(result, Exception):
+                logger.error("Fanout to agent %d failed: %s", i, result)
+                responses.append(None)  # type: ignore[arg-type]
+            else:
+                responses.append(result)  # type: ignore[arg-type]
+
+        logger.info("Fanout pipeline completed for %d agents", len(agents))
+        return responses
+
+    async def shutdown(self, timeout: Optional[float] = 5.0) -> None:
+        """Shutdown all running agents gracefully.
+
+        Args:
+            timeout: Timeout for graceful shutdown
+        """
+        logger.info("Shutting down TeamCoordinator...")
+
+        cancel_tasks = [
+            asyncio.create_task(asyncio.wait_for(task, timeout=timeout))
+            for task in self._running_tasks.values()
+        ]
+
+        if cancel_tasks:
+            await asyncio.gather(*cancel_tasks, return_exceptions=True)
+
+        self._running_tasks.clear()
+        logger.info("TeamCoordinator shutdown complete")
+
+    @property
+    def agents(self) -> Dict[str, Any]:
+        """Get copy of registered agents dict."""
+        return dict(self._agents)
+
+
+__all__ = ["TeamCoordinator"]
--- a/backend/agents/team_pipeline_config.py
+++ b/backend/agents/team_pipeline_config.py
@@ -0,0 +1,132 @@
+# -*- coding: utf-8 -*-
+"""Run-scoped team pipeline configuration helpers."""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Iterable, List, Dict, Any
+
+import yaml
+
+
+DEFAULT_FILENAME = "TEAM_PIPELINE.yaml"
+
+
+def team_pipeline_path(project_root: Path, config_name: str) -> Path:
+    """Return run-scoped team pipeline config path."""
+    return project_root / "runs" / config_name / DEFAULT_FILENAME
+
+
+def ensure_team_pipeline_config(
+    project_root: Path,
+    config_name: str,
+    default_analysts: Iterable[str],
+) -> Path:
+    """Ensure TEAM_PIPELINE.yaml exists for one run."""
+    path = team_pipeline_path(project_root, config_name)
+    path.parent.mkdir(parents=True, exist_ok=True)
+    if path.exists():
+        return path
+
+    payload = {
+        "version": 1,
+        "controller_agent": "portfolio_manager",
+        "discussion": {
+            "allow_dynamic_team_update": True,
+            "active_analysts": list(default_analysts),
+        },
+        "decision": {
+            "require_risk_manager": True,
+        },
+    }
+    path.write_text(
+        yaml.safe_dump(payload, allow_unicode=True, sort_keys=False),
+        encoding="utf-8",
+    )
+    return path
+
+
+def load_team_pipeline_config(project_root: Path, config_name: str) -> Dict[str, Any]:
+    """Load TEAM_PIPELINE.yaml and return parsed dict."""
+    path = team_pipeline_path(project_root, config_name)
+    if not path.exists():
+        return {}
+    parsed = yaml.safe_load(path.read_text(encoding="utf-8")) or {}
+    return parsed if isinstance(parsed, dict) else {}
+
+
+def save_team_pipeline_config(
+    project_root: Path,
+    config_name: str,
+    config: Dict[str, Any],
+) -> Path:
+    """Persist TEAM_PIPELINE.yaml."""
+    path = team_pipeline_path(project_root, config_name)
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(
+        yaml.safe_dump(config, allow_unicode=True, sort_keys=False),
+        encoding="utf-8",
+    )
+    return path
+
+
+def resolve_active_analysts(
+    project_root: Path,
+    config_name: str,
+    available_analysts: Iterable[str],
+) -> List[str]:
+    """Resolve active analysts from TEAM_PIPELINE.yaml."""
+    available = [item for item in available_analysts]
+    parsed = load_team_pipeline_config(project_root, config_name)
+    discussion = parsed.get("discussion", {}) if isinstance(parsed, dict) else {}
+    configured = discussion.get("active_analysts", [])
+    if not isinstance(configured, list) or not configured:
+        return available
+
+    active = [item for item in configured if item in available]
+    return active or available
+
+
+def update_active_analysts(
+    project_root: Path,
+    config_name: str,
+    available_analysts: Iterable[str],
+    *,
+    add: Iterable[str] | None = None,
+    remove: Iterable[str] | None = None,
+    set_to: Iterable[str] | None = None,
+) -> List[str]:
+    """Update active analysts and persist TEAM_PIPELINE.yaml."""
+    available = [item for item in available_analysts]
+    ensure_team_pipeline_config(project_root, config_name, available)
+    parsed = load_team_pipeline_config(project_root, config_name)
+    discussion = parsed.setdefault("discussion", {})
+    if not isinstance(discussion, dict):
+        discussion = {}
+        parsed["discussion"] = discussion
+
+    current = discussion.get("active_analysts", [])
+    if not isinstance(current, list):
+        current = []
+    current = [item for item in current if item in available]
+    if not current:
+        current = list(available)
+
+    if set_to is not None:
+        target = [item for item in set_to if item in available]
+        current = target or current
+
+    for item in add or []:
+        if item in available and item not in current:
+            current.append(item)
+
+    for item in remove or []:
+        current = [existing for existing in current if existing != item]
+
+    if not current:
+        current = [available[0]] if available else []
+
+    discussion["active_analysts"] = current
+    save_team_pipeline_config(project_root, config_name, parsed)
+    return current
+
--- a/backend/agents/templates.py
+++ b/backend/agents/templates.py
@@ -0,0 +1,286 @@
+"""
+Agent模板定义
+
+包含各角色的ROLE.md内容字典，供程序生成Agent工作空间时使用。
+"""
+
+# 基础模板文件内容
+BASE_TEMPLATES = {
+    "AGENTS.md": """# Agent Guide
+
+## 工作流程
+1. 接收分析任务
+2. 调用相关工具/技能
+3. 生成分析报告
+4. 参与团队决策
+
+## 工具使用规范
+- 优先使用已激活的技能
+- 不确定时询问Portfolio Manager
+- 重要发现用 `/save` 记录
+
+## 记忆管理
+- 使用 `/compact` 定期压缩记忆
+- 投资经验记录在MEMORY.md
+""",
+
+    "SOUL.md": """# Soul
+
+你是专业的金融分析师，语气冷静、客观、专业。
+你的分析应该数据驱动，避免情绪化表达。
+""",
+
+    "PROFILE.md": """# Profile
+
+## 投资风格
+- 风险承受能力：中等
+- 投资期限：中期（3-12个月）
+- 偏好行业：科技、医疗、消费
+
+## 优势
+- 财务分析
+- 趋势识别
+
+## 改进方向
+- 市场情绪把握
+""",
+
+    "MEMORY.md": """# Memory
+
+<!-- 此文件用于记录Agent的学习经验和重要发现 -->
+
+## 经验总结
+
+## 重要事件
+
+## 改进记录
+""",
+
+    "HEARTBEAT.md": """# Heartbeat
+
+## 定时任务
+- 每日开盘前检查持仓
+- 收盘后记录当日表现
+""",
+
+    "POLICY.md": """# Policy
+
+## 风控规则
+- 单一持仓不超过20%
+- 止损线：-15%
+""",
+
+    "STYLE.md": """# Style
+
+- 使用结构化输出（JSON/Markdown表格）
+- 包含置信度评分
+- 列出关键假设
+""",
+
+    "agent.yaml": """agent_id: {agent_id}
+agent_type: {agent_type}
+name: {name}
+model:
+  provider: openai
+  model_name: gpt-4o
+  temperature: 0.3
+enabled_skills: []
+disabled_skills: []
+settings: {{}}
+""",
+}
+
+# 角色专用模板
+ROLE_TEMPLATES = {
+    "fundamental": {
+        "ROLE.md": """# Role: Fundamental Analyst
+
+## 职责
+分析公司财务报表、盈利能力、成长性、竞争优势等基本面因素。
+
+## 分析维度
+- 财务报表分析（资产负债表、利润表、现金流量表）
+- 盈利能力指标（ROE、ROA、毛利率、净利率）
+- 成长性指标（营收增长率、利润增长率）
+- 估值指标（P/E、P/B、P/S）
+- 行业地位和竞争优势
+
+## 输出格式
+- 财务健康度评分（1-10）
+- 成长性评分（1-10）
+- 关键财务亮点和风险
+- 同业对比分析
+""",
+        "SOUL.md": """# Soul
+
+你是严谨的基本面分析师，像沃伦·巴菲特一样注重企业内在价值。
+你的分析深入细致，关注长期价值而非短期波动。
+语气沉稳、逻辑严密，善于发现财务数据背后的商业本质。
+""",
+    },
+
+    "technical": {
+        "ROLE.md": """# Role: Technical Analyst
+
+## 职责
+分析价格走势、交易量、技术指标，识别买卖时机。
+
+## 分析维度
+- 趋势分析（长期/中期/短期趋势）
+- 支撑阻力位识别
+- 技术指标（MACD、RSI、KDJ、布林带等）
+- 形态识别（头肩顶/底、双底、三角形等）
+- 量价关系分析
+
+## 输出格式
+- 趋势方向（上涨/下跌/震荡）
+- 关键价位（支撑/阻力）
+- 技术信号（买入/卖出/观望）
+- 置信度评分
+""",
+        "SOUL.md": """# Soul
+
+你是敏锐的技术分析师，相信价格包含一切信息。
+你善于从图表中发现规律，像侦探一样寻找市场留下的痕迹。
+语气果断、快速反应，善于捕捉稍纵即逝的交易机会。
+""",
+    },
+
+    "sentiment": {
+        "ROLE.md": """# Role: Sentiment Analyst
+
+## 职责
+分析市场情绪、资金流向、新闻舆情，判断市场心理状态。
+
+## 分析维度
+- 市场情绪指标（恐慌/贪婪指数）
+- 资金流向分析（主力/散户资金）
+- 新闻舆情分析（正面/负面/中性）
+- 社交媒体情绪
+- 机构持仓变化
+
+## 输出格式
+- 情绪评分（-10到+10，极度恐慌到极度贪婪）
+- 资金流向判断
+- 舆情摘要
+- 情绪拐点预警
+""",
+        "SOUL.md": """# Soul
+
+你是敏感的市场情绪捕手，善于感知市场的恐惧与贪婪。
+你关注人性在金融市场中的表现，理解情绪如何驱动价格。
+语气富有洞察力、善于捕捉微妙变化，像心理学家一样理解市场参与者。
+""",
+    },
+
+    "valuation": {
+        "ROLE.md": """# Role: Valuation Analyst
+
+## 职责
+评估公司内在价值，计算合理价格区间，识别高估/低估机会。
+
+## 分析维度
+- DCF现金流折现模型
+- 相对估值法（P/E、EV/EBITDA等）
+- 资产重估法
+- 分部估值（SOTP）
+- 安全边际计算
+
+## 输出格式
+- 内在价值估算
+- 合理价格区间
+- 当前价格vs内在价值（高估/低估百分比）
+- 估值假设和敏感性分析
+""",
+        "SOUL.md": """# Soul
+
+你是精确的估值分析师，追求计算内在价值的准确区间。
+你像精算师一样严谨，注重假设的合理性和安全边际。
+语气精确、注重数字，善于发现市场定价错误带来的机会。
+""",
+    },
+
+    "portfolio": {
+        "ROLE.md": """# Role: Portfolio Manager
+
+## 职责
+统筹各分析师意见，制定投资决策，管理投资组合配置。
+
+## 分析维度
+- 资产配置策略（股债比例、行业分布）
+- 风险收益平衡
+- 仓位管理（建仓/加仓/减仓/清仓）
+- 再平衡时机
+- 组合相关性分析
+
+## 输出格式
+- 投资决策（买入/卖出/持有）
+- 建议仓位比例
+- 目标价位
+- 止损止盈设置
+- 组合调整建议
+""",
+        "SOUL.md": """# Soul
+
+你是睿智的投资组合经理，像将军一样统筹全局。
+你善于权衡各方意见，做出果断而理性的投资决策。
+语气权威、决策果断，对组合整体表现负有最终责任。
+""",
+    },
+
+    "risk": {
+        "ROLE.md": """# Role: Risk Manager
+
+## 职责
+识别、评估和监控投资风险，确保组合风险在可控范围内。
+
+## 分析维度
+- 市场风险（Beta、波动率）
+- 信用风险
+- 流动性风险
+- 集中度风险
+- 尾部风险（VaR、CVaR）
+- 压力测试
+
+## 输出格式
+- 风险等级（低/中/高/极高）
+- 风险敞口分析
+- 风险调整建议
+- 预警阈值设置
+- 应急预案
+""",
+        "SOUL.md": """# Soul
+
+你是谨慎的风险管理者，时刻警惕潜在的损失。
+你像守门员一样守护组合安全，宁可错过机会也不冒无法承受的风险。
+语气保守、风险意识强，善于发现隐藏的威胁和脆弱性。
+""",
+    },
+}
+
+
+def get_base_template(filename: str) -> str | None:
+    """获取基础模板内容"""
+    return BASE_TEMPLATES.get(filename)
+
+
+def get_role_template(role_type: str, filename: str) -> str | None:
+    """获取角色专用模板内容"""
+    role = ROLE_TEMPLATES.get(role_type)
+    if role:
+        return role.get(filename)
+    return None
+
+
+def get_all_role_types() -> list[str]:
+    """获取所有角色类型列表"""
+    return list(ROLE_TEMPLATES.keys())
+
+
+def render_agent_yaml(agent_id: str, agent_type: str, name: str) -> str:
+    """渲染agent.yaml模板"""
+    return BASE_TEMPLATES["agent.yaml"].format(
+        agent_id=agent_id,
+        agent_type=agent_type,
+        name=name
+    )
--- a/backend/agents/toolkit_factory.py
+++ b/backend/agents/toolkit_factory.py
@@ -0,0 +1,517 @@
+# -*- coding: utf-8 -*-
+"""Toolkit factory following AgentScope's skill + tool group practices.
+
+支持从Agent工作空间动态创建工具集，加载builtin/customized技能，
+以及合并Agent特定工具。
+"""
+
+from typing import Any, Dict, Iterable, List, Optional, Set
+from pathlib import Path
+
+import yaml
+
+from backend.agents.agent_workspace import load_agent_workspace_config
+from backend.agents.skills_manager import SkillsManager
+from backend.agents.skill_loader import load_skill_from_dir, get_skill_tools
+from backend.agents.skill_metadata import parse_skill_metadata
+from backend.config.bootstrap_config import get_bootstrap_config_for_run
+
+
+def load_agent_profiles() -> Dict[str, Dict[str, Any]]:
+    """加载Agent配置文件"""
+    config_path = SkillsManager().project_root / "backend" / "config" / "agent_profiles.yaml"
+    with open(config_path, "r", encoding="utf-8") as file:
+        return yaml.safe_load(file) or {}
+
+
+def _register_analysis_tool_groups(toolkit: Any) -> None:
+    """注册分析工具组"""
+    from backend.tools.analysis_tools import TOOL_REGISTRY
+
+    tool_groups = {
+        "fundamentals": {
+            "description": "Financial health, profitability, growth, and efficiency analysis tools.",
+            "active": False,
+            "notes": (
+                "Use these tools to validate business quality, financial resilience, "
+                "and earnings durability before making directional conclusions."
+            ),
+            "tools": [
+                "analyze_profitability",
+                "analyze_growth",
+                "analyze_financial_health",
+                "analyze_efficiency_ratios",
+                "analyze_valuation_ratios",
+                "get_financial_metrics_tool",
+            ],
+        },
+        "technical": {
+            "description": "Trend, momentum, mean reversion, and volatility analysis tools.",
+            "active": False,
+            "notes": (
+                "Use these tools to assess timing, price structure, and risk-reward in "
+                "the current market regime."
+            ),
+            "tools": [
+                "analyze_trend_following",
+                "analyze_momentum",
+                "analyze_mean_reversion",
+                "analyze_volatility",
+            ],
+        },
+        "sentiment": {
+            "description": "News sentiment and insider activity analysis tools.",
+            "active": False,
+            "notes": (
+                "Use these tools to capture short-horizon catalysts, sentiment shifts, "
+                "and behavioral signals around each ticker."
+            ),
+            "tools": [
+                "analyze_news_sentiment",
+                "analyze_insider_trading",
+            ],
+        },
+        "valuation": {
+            "description": "Intrinsic value and relative valuation analysis tools.",
+            "active": False,
+            "notes": (
+                "Use these tools when the task requires fair value estimation, margin of "
+                "safety analysis, or valuation scenario comparison."
+            ),
+            "tools": [
+                "dcf_valuation_analysis",
+                "owner_earnings_valuation_analysis",
+                "ev_ebitda_valuation_analysis",
+                "residual_income_valuation_analysis",
+            ],
+        },
+    }
+
+    for group_name, group_config in tool_groups.items():
+        toolkit.create_tool_group(
+            group_name=group_name,
+            description=group_config["description"],
+            active=group_config["active"],
+            notes=group_config["notes"],
+        )
+        for tool_name in group_config["tools"]:
+            tool_func = TOOL_REGISTRY.get(tool_name)
+            if tool_func:
+                toolkit.register_tool_function(
+                    tool_func,
+                    group_name=group_name,
+                )
+
+
+def _register_portfolio_tool_groups(toolkit: Any, pm_agent: Any) -> None:
+    """注册投资组合工具组"""
+    toolkit.create_tool_group(
+        group_name="portfolio_ops",
+        description="Portfolio decision recording tools.",
+        active=False,
+        notes=(
+            "Use portfolio tools only after synthesizing analyst and risk inputs. "
+            "Record one explicit decision per ticker."
+        ),
+    )
+    toolkit.register_tool_function(
+        pm_agent._make_decision,
+        group_name="portfolio_ops",
+    )
+    if hasattr(pm_agent, "_add_team_analyst"):
+        toolkit.register_tool_function(
+            pm_agent._add_team_analyst,
+            group_name="portfolio_ops",
+        )
+    if hasattr(pm_agent, "_remove_team_analyst"):
+        toolkit.register_tool_function(
+            pm_agent._remove_team_analyst,
+            group_name="portfolio_ops",
+        )
+    if hasattr(pm_agent, "_set_active_analysts"):
+        toolkit.register_tool_function(
+            pm_agent._set_active_analysts,
+            group_name="portfolio_ops",
+        )
+    if hasattr(pm_agent, "_create_team_analyst"):
+        toolkit.register_tool_function(
+            pm_agent._create_team_analyst,
+            group_name="portfolio_ops",
+        )
+
+
+def _register_risk_tool_groups(toolkit: Any) -> None:
+    """注册风险工具组"""
+    from backend.tools.risk_tools import (
+        assess_margin_and_liquidity,
+        assess_position_concentration,
+        assess_volatility_exposure,
+    )
+
+    toolkit.create_tool_group(
+        group_name="risk_ops",
+        description="Risk diagnostics for concentration, leverage, and volatility.",
+        active=False,
+        notes=(
+            "Use risk tools to quantify concentration, margin pressure, and volatility "
+            "before writing the final risk memo."
+        ),
+    )
+    toolkit.register_tool_function(
+        assess_position_concentration,
+        group_name="risk_ops",
+    )
+    toolkit.register_tool_function(
+        assess_margin_and_liquidity,
+        group_name="risk_ops",
+    )
+    toolkit.register_tool_function(
+        assess_volatility_exposure,
+        group_name="risk_ops",
+    )
+
+
+def create_agent_toolkit(
+    agent_id: str,
+    config_name: str,
+    owner: Any = None,
+    active_skill_dirs: Iterable[str] | None = None,
+) -> Any:
+    """Create a Toolkit with agent skills and grouped tools.
+
+    Args:
+        agent_id: Agent标识符
+        config_name: 运行配置名称
+        owner: Agent实例（用于注册特定方法）
+        active_skill_dirs: 显式指定的活动技能目录列表
+
+    Returns:
+        配置好的Toolkit实例
+    """
+    from agentscope.tool import Toolkit
+
+    profiles = load_agent_profiles()
+    profile = profiles.get(agent_id, {})
+    skills_manager = SkillsManager()
+    agent_config = load_agent_workspace_config(
+        skills_manager.get_agent_asset_dir(config_name, agent_id) / "agent.yaml",
+    )
+    bootstrap_config = get_bootstrap_config_for_run(
+        skills_manager.project_root,
+        config_name,
+    )
+    override = bootstrap_config.agent_override(agent_id)
+    active_groups = override.get(
+        "active_tool_groups",
+        agent_config.active_tool_groups
+        or profile.get("active_tool_groups", []),
+    )
+    disabled_groups = set(agent_config.disabled_tool_groups)
+    if disabled_groups:
+        active_groups = [
+            group_name
+            for group_name in active_groups
+            if group_name not in disabled_groups
+        ]
+
+    toolkit = Toolkit(
+        agent_skill_instruction=(
+            "<system-info>You have access to project skills. Each skill lives in a "
+            "directory and is described by SKILL.md. Follow the skill instructions "
+            "when they are relevant to the current task.</system-info>"
+        ),
+        agent_skill_template="- {name} (dir: {dir}): {description}",
+    )
+
+    if agent_id.endswith("_analyst"):
+        _register_analysis_tool_groups(toolkit)
+    elif agent_id == "portfolio_manager" and owner is not None:
+        _register_portfolio_tool_groups(toolkit, owner)
+    elif agent_id == "risk_manager":
+        _register_risk_tool_groups(toolkit)
+
+    if active_skill_dirs is None:
+        skill_names = skills_manager.resolve_agent_skill_names(
+            config_name=config_name,
+            agent_id=agent_id,
+            default_skills=profile.get("skills", []),
+        )
+        active_skill_dirs = [
+            skills_manager.get_agent_active_root(config_name, agent_id) / skill_name
+            for skill_name in skill_names
+        ]
+
+    for skill_dir in active_skill_dirs:
+        toolkit.register_agent_skill(str(skill_dir))
+
+    apply_skill_tool_restrictions(toolkit, active_skill_dirs)
+
+    if active_groups:
+        toolkit.update_tool_groups(group_names=active_groups, active=True)
+
+    return toolkit
+
+
+def create_toolkit_from_workspace(
+    agent_id: str,
+    config_name: str,
+    owner: Any = None,
+    include_builtin: bool = True,
+    include_customized: bool = True,
+    include_local: bool = True,
+    active_groups: Optional[List[str]] = None,
+) -> Any:
+    """从Agent工作空间创建工具集
+
+    这是create_agent_toolkit的增强版本，支持更灵活的技能加载策略。
+
+    Args:
+        agent_id: Agent标识符
+        config_name: 运行配置名称
+        owner: Agent实例
+        include_builtin: 是否包含builtin技能
+        include_customized: 是否包含customized技能
+        include_local: 是否包含agent-local技能
+        active_groups: 显式指定的活动工具组
+
+    Returns:
+        配置好的Toolkit实例
+    """
+    from agentscope.tool import Toolkit
+
+    skills_manager = SkillsManager()
+    agent_config = load_agent_workspace_config(
+        skills_manager.get_agent_asset_dir(config_name, agent_id) / "agent.yaml",
+    )
+
+    toolkit = Toolkit(
+        agent_skill_instruction=(
+            "<system-info>You have access to project skills. Each skill lives in a "
+            "directory and is described by SKILL.md. Follow the skill instructions "
+            "when they are relevant to the current task.</system-info>"
+        ),
+        agent_skill_template="- {name} (dir: {dir}): {description}",
+    )
+
+    # 注册Agent类型的默认工具组
+    if agent_id.endswith("_analyst"):
+        _register_analysis_tool_groups(toolkit)
+    elif agent_id == "portfolio_manager" and owner is not None:
+        _register_portfolio_tool_groups(toolkit, owner)
+    elif agent_id == "risk_manager":
+        _register_risk_tool_groups(toolkit)
+
+    # 收集所有要加载的技能目录
+    skill_dirs: List[Path] = []
+
+    # 1. 从active目录加载已同步的技能
+    active_root = skills_manager.get_agent_active_root(config_name, agent_id)
+    if active_root.exists():
+        for skill_dir in sorted(active_root.iterdir()):
+            if skill_dir.is_dir() and (skill_dir / "SKILL.md").exists():
+                skill_dirs.append(skill_dir)
+
+    # 2. 从installed目录加载
+    installed_root = skills_manager.get_agent_installed_root(config_name, agent_id)
+    if installed_root.exists():
+        for skill_dir in sorted(installed_root.iterdir()):
+            if skill_dir.is_dir() and (skill_dir / "SKILL.md").exists():
+                if skill_dir not in skill_dirs:
+                    skill_dirs.append(skill_dir)
+
+    # 3. 从local目录加载agent-local技能
+    if include_local:
+        local_root = skills_manager.get_agent_local_root(config_name, agent_id)
+        if local_root.exists():
+            for skill_dir in sorted(local_root.iterdir()):
+                if skill_dir.is_dir() and (skill_dir / "SKILL.md").exists():
+                    if skill_dir not in skill_dirs:
+                        skill_dirs.append(skill_dir)
+
+    # 注册技能到toolkit
+    for skill_dir in skill_dirs:
+        toolkit.register_agent_skill(str(skill_dir))
+
+    apply_skill_tool_restrictions(toolkit, skill_dirs)
+
+    # 激活指定的工具组
+    if active_groups is None:
+        # 从配置中读取
+        profiles = load_agent_profiles()
+        profile = profiles.get(agent_id, {})
+        active_groups = agent_config.active_tool_groups or profile.get("active_tool_groups", [])
+
+        # 应用禁用列表
+        disabled_groups = set(agent_config.disabled_tool_groups)
+        if disabled_groups:
+            active_groups = [g for g in active_groups if g not in disabled_groups]
+
+    if active_groups:
+        toolkit.update_tool_groups(group_names=active_groups, active=True)
+
+    return toolkit
+
+
+def get_toolkit_info(toolkit: Any) -> Dict[str, Any]:
+    """获取工具集信息
+
+    Args:
+        toolkit: Toolkit实例
+
+    Returns:
+        工具集信息字典
+    """
+    info = {
+        "tool_groups": {},
+        "skills": [],
+        "tools_count": 0,
+    }
+
+    # 获取工具组信息
+    groups = getattr(toolkit, "tool_groups", {})
+    for name, group in groups.items():
+        info["tool_groups"][name] = {
+            "description": getattr(group, "description", ""),
+            "active": getattr(group, "active", False),
+            "tools": [t.name for t in getattr(group, "tools", [])],
+        }
+        info["tools_count"] += len(getattr(group, "tools", []))
+
+    # 获取技能信息
+    skills = getattr(toolkit, "agent_skills", [])
+    for skill in skills:
+        info["skills"].append({
+            "name": getattr(skill, "name", "unknown"),
+            "path": getattr(skill, "path", ""),
+            "description": getattr(skill, "description", ""),
+        })
+
+    return info
+
+
+def refresh_toolkit_skills(
+    toolkit: Any,
+    agent_id: str,
+    config_name: str,
+) -> None:
+    """刷新工具集中的技能
+
+    重新从工作空间加载技能，用于运行时技能变更。
+
+    Args:
+        toolkit: Toolkit实例
+        agent_id: Agent标识符
+        config_name: 运行配置名称
+    """
+    skills_manager = SkillsManager()
+
+    # 清除现有技能
+    if hasattr(toolkit, "agent_skills"):
+        toolkit.agent_skills.clear()
+
+    # 重新加载active技能
+    active_root = skills_manager.get_agent_active_root(config_name, agent_id)
+    if active_root.exists():
+        for skill_dir in sorted(active_root.iterdir()):
+            if skill_dir.is_dir() and (skill_dir / "SKILL.md").exists():
+                toolkit.register_agent_skill(str(skill_dir))
+
+    # 重新加载local技能
+    local_root = skills_manager.get_agent_local_root(config_name, agent_id)
+    if local_root.exists():
+        for skill_dir in sorted(local_root.iterdir()):
+            if skill_dir.is_dir() and (skill_dir / "SKILL.md").exists():
+                toolkit.register_agent_skill(str(skill_dir))
+
+
+def apply_skill_tool_restrictions(toolkit: Any, skill_dirs: List[Path]) -> None:
+    """Apply per-skill allowed_tools / denied_tools restrictions to a toolkit.
+
+    If a skill specifies allowed_tools, only those tools are accessible when
+    that skill is active. If a skill specifies denied_tools, those tools are
+    removed regardless of allowed_tools. Denied tools take precedence.
+
+    This function annotates the toolkit with a _skill_tool_restrictions map
+    that downstream code can consult when resolving available tools.
+
+    Args:
+        toolkit: The agentscope Toolkit instance.
+        skill_dirs: List of skill directory paths to inspect.
+    """
+    restrictions: Dict[str, Dict[str, Set[str]]] = {}
+    for skill_dir in skill_dirs:
+        metadata = parse_skill_metadata(skill_dir, source="active")
+        if not metadata.allowed_tools and not metadata.denied_tools:
+            continue
+        restrictions[skill_dir.name] = {
+            "allowed": set(metadata.allowed_tools),
+            "denied": set(metadata.denied_tools),
+        }
+    if hasattr(toolkit, "agent_skills"):
+        for skill in toolkit.agent_skills:
+            skill_name = getattr(skill, "name", "") or ""
+            if skill_name in restrictions:
+                setattr(
+                    skill,
+                    "_tool_allowed",
+                    restrictions[skill_name]["allowed"],
+                )
+                setattr(
+                    skill,
+                    "_tool_denied",
+                    restrictions[skill_name]["denied"],
+                )
+
+
+def get_skill_effective_tools(skill: Any) -> Optional[Set[str]]:
+    """Return the effective tool set for a skill after applying restrictions.
+
+    If the skill has no restrictions (no allowed_tools / denied_tools),
+    returns None to indicate "all tools allowed".
+
+    If allowed_tools is set, returns only those tools minus denied_tools.
+    If only denied_tools is set, returns all tools minus denied_tools.
+
+    Args:
+        skill: A skill object previously registered via register_agent_skill.
+
+    Returns:
+        A set of allowed tool names, or None if unrestricted.
+    """
+    allowed = getattr(skill, "_tool_allowed", None)
+    denied = getattr(skill, "_tool_denied", set())
+
+    if allowed is None:
+        return None
+
+    effective = allowed - denied
+    return effective
+
+
+def filter_toolkit_by_skill(
+    toolkit: Any,
+    skill_name: str,
+) -> Set[str]:
+    """Return the set of tool names that are accessible for a given skill.
+
+    Args:
+        toolkit: The agentscope Toolkit instance.
+        skill_name: Name of the skill to query.
+
+    Returns:
+        Set of allowed tool names, or all registered tool names if unrestricted.
+    """
+    if not hasattr(toolkit, "agent_skills"):
+        return set()
+
+    for skill in toolkit.agent_skills:
+        name = getattr(skill, "name", "") or ""
+        if name != skill_name:
+            continue
+        effective = get_skill_effective_tools(skill)
+        if effective is None:
+            return set()
+        return effective
+
+    return set()
+
--- a/backend/agents/workspace.py
+++ b/backend/agents/workspace.py
@@ -0,0 +1,327 @@
+# -*- coding: utf-8 -*-
+"""Workspace Manager - Create and manage agent workspaces."""
+
+import logging
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+import yaml
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class WorkspaceConfig:
+    """Configuration for a workspace."""
+
+    workspace_id: str
+    name: str = ""
+    description: str = ""
+    created_at: str = ""
+    metadata: Dict[str, Any] = field(default_factory=dict)
+
+    def to_dict(self) -> Dict[str, Any]:
+        """Serialize to dictionary."""
+        return {
+            "workspace_id": self.workspace_id,
+            "name": self.name,
+            "description": self.description,
+            "created_at": self.created_at,
+            "metadata": self.metadata,
+        }
+
+    @classmethod
+    def from_dict(cls, data: Dict[str, Any]) -> "WorkspaceConfig":
+        """Create from dictionary."""
+        return cls(
+            workspace_id=data.get("workspace_id", ""),
+            name=data.get("name", ""),
+            description=data.get("description", ""),
+            created_at=data.get("created_at", ""),
+            metadata=data.get("metadata", {}),
+        )
+
+
+class WorkspaceRegistry:
+    """Registry for persistent workspace definitions (design-time)."""
+
+    def __init__(self, project_root: Optional[Path] = None):
+        """Initialize the workspace manager.
+
+        Args:
+            project_root: Root directory of the project
+        """
+        self.project_root = project_root or Path(__file__).parent.parent.parent
+        self.workspaces_root = self.project_root / "workspaces"
+        self.workspaces_root.mkdir(parents=True, exist_ok=True)
+
+    def create_workspace(
+        self,
+        workspace_id: str,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        metadata: Optional[Dict[str, Any]] = None,
+    ) -> WorkspaceConfig:
+        """Create a new workspace with directory structure.
+
+        Args:
+            workspace_id: Unique identifier for the workspace
+            name: Display name for the workspace
+            description: Optional description
+            metadata: Optional metadata dictionary
+
+        Returns:
+            WorkspaceConfig instance
+
+        Raises:
+            ValueError: If workspace already exists
+        """
+        workspace_dir = self.workspaces_root / workspace_id
+
+        if workspace_dir.exists():
+            raise ValueError(f"Workspace '{workspace_id}' already exists")
+
+        # Create directory structure
+        workspace_dir.mkdir(parents=True, exist_ok=True)
+
+        # Create subdirectories
+        (workspace_dir / "agents").mkdir(exist_ok=True)
+        (workspace_dir / "shared" / "market_data").mkdir(parents=True, exist_ok=True)
+        (workspace_dir / "shared" / "memories").mkdir(parents=True, exist_ok=True)
+
+        # Create workspace.yaml
+        from datetime import datetime
+
+        config = WorkspaceConfig(
+            workspace_id=workspace_id,
+            name=name or workspace_id,
+            description=description or "",
+            created_at=datetime.now().isoformat(),
+            metadata=metadata or {},
+        )
+
+        self._write_workspace_config(workspace_dir, config)
+
+        return config
+
+    def list_workspaces(self) -> List[WorkspaceConfig]:
+        """List all workspaces.
+
+        Returns:
+            List of WorkspaceConfig instances
+        """
+        workspaces = []
+
+        if not self.workspaces_root.exists():
+            return workspaces
+
+        for workspace_dir in self.workspaces_root.iterdir():
+            if not workspace_dir.is_dir():
+                continue
+
+            config_path = workspace_dir / "workspace.yaml"
+            if config_path.exists():
+                try:
+                    with open(config_path, "r", encoding="utf-8") as f:
+                        data = yaml.safe_load(f) or {}
+                    workspaces.append(WorkspaceConfig.from_dict(data))
+                except Exception as e:
+                    logger.warning(f"Failed to load workspace config {config_path}: {e}")
+
+        return workspaces
+
+    def get_workspace_agents(self, workspace_id: str) -> List[Dict[str, Any]]:
+        """Get all agents in a workspace.
+
+        Args:
+            workspace_id: ID of the workspace
+
+        Returns:
+            List of agent information dictionaries
+
+        Raises:
+            ValueError: If workspace doesn't exist
+        """
+        workspace_dir = self.workspaces_root / workspace_id
+
+        if not workspace_dir.exists():
+            raise ValueError(f"Workspace '{workspace_id}' does not exist")
+
+        agents = []
+        agents_dir = workspace_dir / "agents"
+
+        if not agents_dir.exists():
+            return agents
+
+        for agent_dir in agents_dir.iterdir():
+            if not agent_dir.is_dir():
+                continue
+
+            config_path = agent_dir / "agent.yaml"
+            if config_path.exists():
+                try:
+                    with open(config_path, "r", encoding="utf-8") as f:
+                        config = yaml.safe_load(f) or {}
+
+                    agents.append({
+                        "agent_id": agent_dir.name,
+                        "agent_type": config.get("agent_type", "unknown"),
+                        "config_path": str(config_path),
+                    })
+                except Exception as e:
+                    logger.warning(f"Failed to load agent config {config_path}: {e}")
+
+        return agents
+
+    def get_agent_workspace(self, agent_id: str, workspace_id: str) -> Optional[Path]:
+        """Get the workspace path for an agent.
+
+        Args:
+            agent_id: ID of the agent
+            workspace_id: ID of the workspace
+
+        Returns:
+            Path to agent directory, or None if not found
+        """
+        agent_dir = self.workspaces_root / workspace_id / "agents" / agent_id
+
+        if agent_dir.exists():
+            return agent_dir
+
+        return None
+
+    def workspace_exists(self, workspace_id: str) -> bool:
+        """Check if a workspace exists.
+
+        Args:
+            workspace_id: ID of the workspace
+
+        Returns:
+            True if workspace exists, False otherwise
+        """
+        workspace_dir = self.workspaces_root / workspace_id
+        return workspace_dir.exists() and (workspace_dir / "workspace.yaml").exists()
+
+    def delete_workspace(self, workspace_id: str, force: bool = False) -> bool:
+        """Delete a workspace and all its agents.
+
+        Args:
+            workspace_id: ID of the workspace to delete
+            force: If True, delete even if workspace has agents
+
+        Returns:
+            True if deleted, False if workspace didn't exist
+
+        Raises:
+            ValueError: If workspace has agents and force is False
+        """
+        import shutil
+
+        workspace_dir = self.workspaces_root / workspace_id
+
+        if not workspace_dir.exists():
+            return False
+
+        # Check for agents
+        agents_dir = workspace_dir / "agents"
+        if agents_dir.exists() and any(agents_dir.iterdir()):
+            if not force:
+                raise ValueError(
+                    f"Workspace '{workspace_id}' contains agents. "
+                    "Use force=True to delete anyway."
+                )
+
+        shutil.rmtree(workspace_dir)
+        return True
+
+    def get_workspace_path(self, workspace_id: str) -> Path:
+        """Get the path to a workspace directory.
+
+        Args:
+            workspace_id: ID of the workspace
+
+        Returns:
+            Path to workspace directory
+        """
+        return self.workspaces_root / workspace_id
+
+    def get_shared_data_path(self, workspace_id: str) -> Optional[Path]:
+        """Get the shared data directory for a workspace.
+
+        Args:
+            workspace_id: ID of the workspace
+
+        Returns:
+            Path to shared data directory, or None if workspace doesn't exist
+        """
+        workspace_dir = self.workspaces_root / workspace_id
+
+        if not workspace_dir.exists():
+            return None
+
+        return workspace_dir / "shared"
+
+    def update_workspace_config(
+        self,
+        workspace_id: str,
+        name: Optional[str] = None,
+        description: Optional[str] = None,
+        metadata: Optional[Dict[str, Any]] = None,
+    ) -> WorkspaceConfig:
+        """Update workspace configuration.
+
+        Args:
+            workspace_id: ID of the workspace
+            name: New display name (optional)
+            description: New description (optional)
+            metadata: Metadata to merge (optional)
+
+        Returns:
+            Updated WorkspaceConfig
+
+        Raises:
+            ValueError: If workspace doesn't exist
+        """
+        workspace_dir = self.workspaces_root / workspace_id
+
+        if not workspace_dir.exists():
+            raise ValueError(f"Workspace '{workspace_id}' does not exist")
+
+        config_path = workspace_dir / "workspace.yaml"
+        current_config = {}
+
+        if config_path.exists():
+            try:
+                with open(config_path, "r", encoding="utf-8") as f:
+                    current_config = yaml.safe_load(f) or {}
+            except Exception as e:
+                logger.warning(f"Failed to load existing config {config_path}: {e}")
+
+        # Update fields
+        if name is not None:
+            current_config["name"] = name
+        if description is not None:
+            current_config["description"] = description
+        if metadata is not None:
+            current_config["metadata"] = {**current_config.get("metadata", {}), **metadata}
+
+        config = WorkspaceConfig.from_dict(current_config)
+        self._write_workspace_config(workspace_dir, config)
+
+        return config
+
+    def _write_workspace_config(self, workspace_dir: Path, config: WorkspaceConfig) -> None:
+        """Write workspace configuration to file.
+
+        Args:
+            workspace_dir: Workspace directory
+            config: Workspace configuration
+        """
+        config_path = workspace_dir / "workspace.yaml"
+        with open(config_path, "w", encoding="utf-8") as f:
+            yaml.safe_dump(config.to_dict(), f, allow_unicode=True, sort_keys=False)
+
+
+# Backward-compatible alias: legacy imports expect WorkspaceManager.
+WorkspaceManager = WorkspaceRegistry
--- a/backend/agents/workspace_manager.py
+++ b/backend/agents/workspace_manager.py
@@ -0,0 +1,214 @@
+# -*- coding: utf-8 -*-
+"""Initialize run-scoped agent workspace assets."""
+
+from pathlib import Path
+from typing import Dict, Iterable, Optional
+
+import yaml
+
+from .skills_manager import SkillsManager
+from .team_pipeline_config import ensure_team_pipeline_config
+
+
+class RunWorkspaceManager:
+    """Create and maintain run-level prompt asset files for each agent."""
+
+    def __init__(self, project_root: Optional[Path] = None):
+        self.skills_manager = SkillsManager(project_root=project_root)
+        self.project_root = self.skills_manager.project_root
+
+    def get_run_dir(self, config_name: str) -> Path:
+        return self.project_root / "runs" / config_name
+
+    def ensure_run_workspace(self, config_name: str) -> Path:
+        run_dir = self.get_run_dir(config_name)
+        run_dir.mkdir(parents=True, exist_ok=True)
+        self.skills_manager.ensure_activation_manifest(config_name)
+        ensure_team_pipeline_config(
+            project_root=self.project_root,
+            config_name=config_name,
+            default_analysts=[
+                "fundamentals_analyst",
+                "technical_analyst",
+                "sentiment_analyst",
+                "valuation_analyst",
+            ],
+        )
+        bootstrap_path = run_dir / "BOOTSTRAP.md"
+        if not bootstrap_path.exists():
+            bootstrap_path.write_text(
+                "---\n"
+                "tickers:\n"
+                "  - AAPL\n"
+                "  - MSFT\n"
+                "initial_cash: 100000\n"
+                "margin_requirement: 0.0\n"
+                "enable_memory: false\n"
+                "max_comm_cycles: 2\n"
+                "agent_overrides: {}\n"
+                "---\n\n"
+                "# Bootstrap\n\n"
+                "Use this file to describe run-specific setup notes, preferred tickers,\n"
+                "risk bounds, or strategy constraints before the first execution.\n\n"
+                "The YAML front matter above is machine-readable runtime configuration.\n"
+                "The markdown body below is injected into agent prompts as run context.\n",
+                encoding="utf-8",
+            )
+        return run_dir
+
+    def bootstrap_path(self, config_name: str) -> Path:
+        return self.get_run_dir(config_name) / "BOOTSTRAP.md"
+
+    def ensure_agent_assets(
+        self,
+        config_name: str,
+        agent_id: str,
+        role_seed: str = "",
+        style_seed: str = "",
+        policy_seed: str = "",
+    ) -> Path:
+        asset_dir = self.skills_manager.get_agent_asset_dir(
+            config_name,
+            agent_id,
+        )
+        asset_dir.mkdir(parents=True, exist_ok=True)
+        (asset_dir / "skills" / "installed").mkdir(parents=True, exist_ok=True)
+        (asset_dir / "skills" / "active").mkdir(parents=True, exist_ok=True)
+        (asset_dir / "skills" / "disabled").mkdir(parents=True, exist_ok=True)
+        (asset_dir / "skills" / "local").mkdir(parents=True, exist_ok=True)
+
+        self._ensure_file(
+            asset_dir / "ROLE.md",
+            "# Role\n\n"
+            "Optional run-scoped role override.\n\n"
+            f"{role_seed}".strip()
+            + "\n",
+        )
+        self._ensure_file(
+            asset_dir / "STYLE.md",
+            "# Style\n\n"
+            "Optional run-scoped communication or reasoning style.\n\n"
+            f"{style_seed}".strip()
+            + "\n",
+        )
+        self._ensure_file(
+            asset_dir / "POLICY.md",
+            "# Policy\n\n"
+            "Optional run-scoped constraints, limits, or strategy policy.\n\n"
+            f"{policy_seed}".strip()
+            + "\n",
+        )
+        self._ensure_file(
+            asset_dir / "SOUL.md",
+            "# Soul\n\n"
+            "Describe the agent's temperament, reasoning posture, and voice.\n\n",
+        )
+        self._ensure_file(
+            asset_dir / "PROFILE.md",
+            "# Profile\n\n"
+            "Track this agent's long-lived investment style, preferences, and strengths.\n\n",
+        )
+        self._ensure_file(
+            asset_dir / "AGENTS.md",
+            "# Agent Guide\n\n"
+            "Document how this agent should work, collaborate, and choose tools or skills.\n\n",
+        )
+        self._ensure_file(
+            asset_dir / "MEMORY.md",
+            "# Memory\n\n"
+            "Store durable lessons, heuristics, and reminders for this agent.\n\n",
+        )
+        self._ensure_file(
+            asset_dir / "HEARTBEAT.md",
+            "# Heartbeat\n\n"
+            "Optional checklist for periodic review or self-reflection.\n\n",
+        )
+        self._ensure_agent_yaml(
+            asset_dir / "agent.yaml",
+            agent_id=agent_id,
+        )
+        return asset_dir
+
+    def initialize_default_assets(
+        self,
+        config_name: str,
+        agent_ids: Iterable[str],
+        analyst_personas: Optional[Dict[str, Dict]] = None,
+    ) -> None:
+        self.ensure_run_workspace(config_name)
+        analyst_personas = analyst_personas or {}
+
+        for agent_id in agent_ids:
+            if agent_id.endswith("_analyst"):
+                persona = analyst_personas.get(agent_id, {})
+                role_seed = persona.get("description", "").strip()
+                focus_items = persona.get("focus", [])
+                style_seed = "\n".join(f"- {item}" for item in focus_items)
+                policy_seed = (
+                    "State a clear signal, confidence, and the conditions that would invalidate the thesis."
+                )
+            elif agent_id == "portfolio_manager":
+                role_seed = (
+                    "Synthesize analyst and risk inputs into explicit portfolio decisions."
+                )
+                style_seed = (
+                    "Be concise, capital-aware, and explicit about sizing rationale."
+                )
+                policy_seed = (
+                    "Respect cash, margin, and portfolio concentration constraints before recording decisions."
+                )
+            elif agent_id == "risk_manager":
+                role_seed = (
+                    "Quantify concentration, leverage, liquidity, and volatility risk before trade execution."
+                )
+                style_seed = (
+                    "Prioritize the highest-severity risk first and state concrete limits."
+                )
+                policy_seed = (
+                    "Use available risk tools before issuing the final risk memo."
+                )
+            else:
+                role_seed = ""
+                style_seed = ""
+                policy_seed = ""
+
+            self.ensure_agent_assets(
+                config_name=config_name,
+                agent_id=agent_id,
+                role_seed=role_seed,
+                style_seed=style_seed,
+                policy_seed=policy_seed,
+            )
+
+    @staticmethod
+    def _ensure_file(path: Path, content: str) -> None:
+        if not path.exists():
+            path.write_text(content, encoding="utf-8")
+
+    @staticmethod
+    def _ensure_agent_yaml(path: Path, agent_id: str) -> None:
+        if path.exists():
+            return
+
+        payload = {
+            "agent_id": agent_id,
+            "prompt_files": [
+                "SOUL.md",
+                "PROFILE.md",
+                "AGENTS.md",
+                "POLICY.md",
+                "MEMORY.md",
+            ],
+            "enabled_skills": [],
+            "disabled_skills": [],
+            "active_tool_groups": [],
+            "disabled_tool_groups": [],
+        }
+        path.write_text(
+            yaml.safe_dump(payload, allow_unicode=True, sort_keys=False),
+            encoding="utf-8",
+        )
+
+
+# Backward-compatible alias: code importing WorkspaceManager from this module should continue to work.
+WorkspaceManager = RunWorkspaceManager
--- a/backend/api/init.py
+++ b/backend/api/init.py
@@ -0,0 +1,21 @@
+# -*- coding: utf-8 -*-
+"""
+API Routes Package
+
+Provides REST API endpoints for:
+- Agent management
+- Workspace management
+- Tool guard operations
+"""
+
+from .agents import router as agents_router
+from .workspaces import router as workspaces_router
+from .guard import router as guard_router
+from .runtime import router as runtime_router
+
+__all__ = [
+    "agents_router",
+    "workspaces_router",
+    "guard_router",
+    "runtime_router",
+]
--- a/backend/api/agents.py
+++ b/backend/api/agents.py
@@ -0,0 +1,497 @@
+# -*- coding: utf-8 -*-
+"""
+Agent API Routes
+
+Provides REST API endpoints for agent management within workspaces.
+"""
+import logging
+import os
+import tempfile
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+from fastapi import APIRouter, HTTPException, Depends, Body, UploadFile, File, Form
+from pydantic import BaseModel, Field
+
+from backend.agents import AgentFactory, WorkspaceManager, get_registry
+from backend.agents.skills_manager import SkillsManager
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter(prefix="/api/workspaces/{workspace_id}/agents", tags=["agents"])
+
+
+# Request/Response Models
+class CreateAgentRequest(BaseModel):
+    """Request to create a new agent."""
+    agent_id: str = Field(..., description="Unique agent identifier")
+    agent_type: str = Field(..., description="Type of agent (e.g., technical_analyst)")
+    name: Optional[str] = Field(None, description="Display name")
+    description: Optional[str] = Field(None, description="Agent description")
+    clone_from: Optional[str] = Field(None, description="Agent ID to clone from")
+    llm_model_config: Optional[Dict[str, Any]] = Field(None, description="LLM model configuration")
+
+
+class UpdateAgentRequest(BaseModel):
+    """Request to update an agent."""
+    name: Optional[str] = None
+    description: Optional[str] = None
+    enabled_skills: Optional[List[str]] = None
+    disabled_skills: Optional[List[str]] = None
+
+
+class InstallExternalSkillRequest(BaseModel):
+    """Request to install an external skill for one agent."""
+    source: str = Field(..., description="Directory path, zip path, or http(s) zip URL")
+    name: Optional[str] = Field(None, description="Optional override skill name")
+    activate: bool = Field(True, description="Whether to enable skill immediately")
+
+
+class AgentResponse(BaseModel):
+    """Agent information response."""
+    agent_id: str
+    agent_type: str
+    workspace_id: str
+    config_path: str
+    agent_dir: str
+    status: str = "inactive"
+
+
+class AgentFileResponse(BaseModel):
+    """Agent file content response."""
+    filename: str
+    content: str
+
+
+# Dependencies
+def get_agent_factory():
+    """Get AgentFactory instance."""
+    return AgentFactory()
+
+
+def get_workspace_manager():
+    """Get WorkspaceManager instance."""
+    return WorkspaceManager()
+
+
+def get_skills_manager():
+    """Get SkillsManager instance."""
+    return SkillsManager()
+
+
+# Routes
+@router.post("", response_model=AgentResponse)
+async def create_agent(
+    workspace_id: str,
+    request: CreateAgentRequest,
+    factory: AgentFactory = Depends(get_agent_factory),
+    registry = Depends(get_registry),
+):
+    """
+    Create a new agent in a workspace.
+
+    Args:
+        workspace_id: Workspace identifier
+        request: Agent creation parameters
+
+    Returns:
+        Created agent information
+    """
+    # Check workspace exists
+    if not factory.workspaces_root.exists():
+        raise HTTPException(status_code=404, detail="Workspaces root not found")
+
+    workspace_dir = factory.workspaces_root / workspace_id
+    if not workspace_dir.exists():
+        raise HTTPException(status_code=404, detail=f"Workspace '{workspace_id}' not found")
+
+    try:
+        # Create agent
+        agent = factory.create_agent(
+            agent_id=request.agent_id,
+            agent_type=request.agent_type,
+            workspace_id=workspace_id,
+            clone_from=request.clone_from,
+        )
+
+        # Register in registry
+        registry.register(
+            agent_id=request.agent_id,
+            agent_type=request.agent_type,
+            workspace_id=workspace_id,
+            config_path=str(agent.config_path),
+            agent_dir=str(agent.agent_dir),
+            status="inactive",
+        )
+
+        return AgentResponse(
+            agent_id=agent.agent_id,
+            agent_type=agent.agent_type,
+            workspace_id=agent.workspace_id,
+            config_path=str(agent.config_path),
+            agent_dir=str(agent.agent_dir),
+            status="inactive",
+        )
+
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+
+@router.get("", response_model=List[AgentResponse])
+async def list_agents(
+    workspace_id: str,
+    factory: AgentFactory = Depends(get_agent_factory),
+):
+    """
+    List all agents in a workspace.
+
+    Args:
+        workspace_id: Workspace identifier
+
+    Returns:
+        List of agents
+    """
+    try:
+        agents_data = factory.list_agents(workspace_id=workspace_id)
+        return [
+            AgentResponse(
+                agent_id=agent["agent_id"],
+                agent_type=agent["agent_type"],
+                workspace_id=workspace_id,
+                config_path=agent["config_path"],
+                agent_dir=str(Path(agent["config_path"]).parent),
+                status="inactive",
+            )
+            for agent in agents_data
+        ]
+    except ValueError as e:
+        raise HTTPException(status_code=404, detail=str(e))
+
+
+@router.get("/{agent_id}", response_model=AgentResponse)
+async def get_agent(
+    workspace_id: str,
+    agent_id: str,
+    registry = Depends(get_registry),
+):
+    """
+    Get agent details.
+
+    Args:
+        workspace_id: Workspace identifier
+        agent_id: Agent identifier
+
+    Returns:
+        Agent information
+    """
+    agent_info = registry.get(agent_id)
+
+    if not agent_info or agent_info.workspace_id != workspace_id:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    return AgentResponse(
+        agent_id=agent_info.agent_id,
+        agent_type=agent_info.agent_type,
+        workspace_id=agent_info.workspace_id,
+        config_path=agent_info.config_path,
+        agent_dir=agent_info.agent_dir,
+        status=agent_info.status,
+    )
+
+
+@router.delete("/{agent_id}")
+async def delete_agent(
+    workspace_id: str,
+    agent_id: str,
+    factory: AgentFactory = Depends(get_agent_factory),
+    registry = Depends(get_registry),
+):
+    """
+    Delete an agent.
+
+    Args:
+        workspace_id: Workspace identifier
+        agent_id: Agent identifier
+
+    Returns:
+        Success message
+    """
+    # Check agent exists in registry
+    agent_info = registry.get(agent_id)
+    if not agent_info or agent_info.workspace_id != workspace_id:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    # Delete from factory
+    success = factory.delete_agent(agent_id, workspace_id)
+    if not success:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    # Unregister
+    registry.unregister(agent_id)
+
+    return {"message": f"Agent '{agent_id}' deleted successfully"}
+
+
+@router.patch("/{agent_id}", response_model=AgentResponse)
+async def update_agent(
+    workspace_id: str,
+    agent_id: str,
+    request: UpdateAgentRequest,
+    registry = Depends(get_registry),
+):
+    """
+    Update agent configuration.
+
+    Args:
+        workspace_id: Workspace identifier
+        agent_id: Agent identifier
+        request: Update parameters
+
+    Returns:
+        Updated agent information
+    """
+    agent_info = registry.get(agent_id)
+    if not agent_info or agent_info.workspace_id != workspace_id:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    # Update metadata in registry
+    metadata_updates = {}
+    if request.name:
+        metadata_updates["name"] = request.name
+    if request.description:
+        metadata_updates["description"] = request.description
+
+    if metadata_updates:
+        registry.update_metadata(agent_id, metadata_updates)
+
+    # Update skills if provided
+    if request.enabled_skills or request.disabled_skills:
+        skills_manager = SkillsManager()
+        skills_manager.update_agent_skill_overrides(
+            config_name=workspace_id,
+            agent_id=agent_id,
+            enable=request.enabled_skills or [],
+            disable=request.disabled_skills or [],
+        )
+
+    # Get updated info
+    agent_info = registry.get(agent_id)
+    return AgentResponse(
+        agent_id=agent_info.agent_id,
+        agent_type=agent_info.agent_type,
+        workspace_id=agent_info.workspace_id,
+        config_path=agent_info.config_path,
+        agent_dir=agent_info.agent_dir,
+        status=agent_info.status,
+    )
+
+
+@router.post("/{agent_id}/skills/{skill_name}/enable")
+async def enable_skill(
+    workspace_id: str,
+    agent_id: str,
+    skill_name: str,
+    registry = Depends(get_registry),
+):
+    """
+    Enable a skill for an agent.
+
+    Args:
+        workspace_id: Workspace identifier
+        agent_id: Agent identifier
+        skill_name: Skill name to enable
+
+    Returns:
+        Success message
+    """
+    agent_info = registry.get(agent_id)
+    if not agent_info or agent_info.workspace_id != workspace_id:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    skills_manager = SkillsManager()
+    result = skills_manager.update_agent_skill_overrides(
+        config_name=workspace_id,
+        agent_id=agent_id,
+        enable=[skill_name],
+    )
+
+    return {
+        "message": f"Skill '{skill_name}' enabled for agent '{agent_id}'",
+        "enabled_skills": result["enabled_skills"],
+    }
+
+
+@router.post("/{agent_id}/skills/{skill_name}/disable")
+async def disable_skill(
+    workspace_id: str,
+    agent_id: str,
+    skill_name: str,
+    registry = Depends(get_registry),
+):
+    """
+    Disable a skill for an agent.
+
+    Args:
+        workspace_id: Workspace identifier
+        agent_id: Agent identifier
+        skill_name: Skill name to disable
+
+    Returns:
+        Success message
+    """
+    agent_info = registry.get(agent_id)
+    if not agent_info or agent_info.workspace_id != workspace_id:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    skills_manager = SkillsManager()
+    result = skills_manager.update_agent_skill_overrides(
+        config_name=workspace_id,
+        agent_id=agent_id,
+        disable=[skill_name],
+    )
+
+    return {
+        "message": f"Skill '{skill_name}' disabled for agent '{agent_id}'",
+        "disabled_skills": result["disabled_skills"],
+    }
+
+
+@router.post("/{agent_id}/skills/install")
+async def install_external_skill(
+    workspace_id: str,
+    agent_id: str,
+    request: InstallExternalSkillRequest,
+    registry=Depends(get_registry),
+):
+    """Install an external skill into one agent's local skills."""
+    agent_info = registry.get(agent_id)
+    if not agent_info or agent_info.workspace_id != workspace_id:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    skills_manager = SkillsManager()
+    try:
+        result = skills_manager.install_external_skill_for_agent(
+            config_name=workspace_id,
+            agent_id=agent_id,
+            source=request.source,
+            skill_name=request.name,
+            activate=request.activate,
+        )
+    except (FileNotFoundError, ValueError) as exc:
+        raise HTTPException(status_code=400, detail=str(exc))
+
+    return {
+        "message": f"Installed external skill '{result['skill_name']}' for '{agent_id}'",
+        **result,
+    }
+
+
+@router.post("/{agent_id}/skills/upload")
+async def upload_external_skill(
+    workspace_id: str,
+    agent_id: str,
+    file: UploadFile = File(...),
+    name: Optional[str] = Form(None),
+    activate: bool = Form(True),
+    registry=Depends(get_registry),
+):
+    """Upload a zip skill package from frontend and install for one agent."""
+    agent_info = registry.get(agent_id)
+    if not agent_info or agent_info.workspace_id != workspace_id:
+        raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+
+    original_name = (file.filename or "").strip()
+    if not original_name.lower().endswith(".zip"):
+        raise HTTPException(status_code=400, detail="Uploaded file must be a .zip archive")
+
+    suffix = Path(original_name).suffix or ".zip"
+    temp_path: Optional[str] = None
+    try:
+        with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as tmp:
+            temp_path = tmp.name
+            content = await file.read()
+            tmp.write(content)
+
+        skills_manager = SkillsManager()
+        result = skills_manager.install_external_skill_for_agent(
+            config_name=workspace_id,
+            agent_id=agent_id,
+            source=temp_path,
+            skill_name=name,
+            activate=activate,
+        )
+    except (FileNotFoundError, ValueError) as exc:
+        raise HTTPException(status_code=400, detail=str(exc))
+    finally:
+        try:
+            await file.close()
+        except Exception as e:
+            logger.warning(f"Failed to close uploaded file: {e}")
+        if temp_path and os.path.exists(temp_path):
+            os.remove(temp_path)
+
+    return {
+        "message": f"Uploaded and installed external skill '{result['skill_name']}' for '{agent_id}'",
+        **result,
+    }
+
+
+@router.get("/{agent_id}/files/{filename}", response_model=AgentFileResponse)
+async def get_agent_file(
+    workspace_id: str,
+    agent_id: str,
+    filename: str,
+    workspace_manager: WorkspaceManager = Depends(get_workspace_manager),
+):
+    """
+    Read an agent's workspace file.
+
+    Args:
+        workspace_id: Workspace identifier
+        agent_id: Agent identifier
+        filename: File to read (e.g., SOUL.md, ROLE.md)
+
+    Returns:
+        File content
+    """
+    try:
+        content = workspace_manager.load_agent_file(
+            config_name=workspace_id,
+            agent_id=agent_id,
+            filename=filename,
+        )
+        return AgentFileResponse(filename=filename, content=content)
+    except FileNotFoundError:
+        raise HTTPException(status_code=404, detail=f"File '{filename}' not found")
+
+
+@router.put("/{agent_id}/files/{filename}", response_model=AgentFileResponse)
+async def update_agent_file(
+    workspace_id: str,
+    agent_id: str,
+    filename: str,
+    content: str = Body(..., media_type="text/plain"),
+    workspace_manager: WorkspaceManager = Depends(get_workspace_manager),
+):
+    """
+    Update an agent's workspace file.
+
+    Args:
+        workspace_id: Workspace identifier
+        agent_id: Agent identifier
+        filename: File to update
+        content: New file content
+
+    Returns:
+        Updated file information
+    """
+    try:
+        workspace_manager.update_agent_file(
+            config_name=workspace_id,
+            agent_id=agent_id,
+            filename=filename,
+            content=content,
+        )
+        return AgentFileResponse(filename=filename, content=content)
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
--- a/backend/api/guard.py
+++ b/backend/api/guard.py
@@ -0,0 +1,257 @@
+# -*- coding: utf-8 -*-
+"""
+Tool Guard API Routes
+
+Provides REST API endpoints for tool guard operations.
+"""
+from __future__ import annotations
+
+from typing import Any, Dict, List, Optional
+from datetime import datetime
+
+from fastapi import APIRouter, HTTPException
+from pydantic import BaseModel, Field
+
+from backend.agents.base.tool_guard import (
+    ApprovalRecord,
+    ApprovalStatus,
+    SeverityLevel,
+    TOOL_GUARD_STORE,
+    default_findings_for_tool,
+)
+
+router = APIRouter(prefix="/api/guard", tags=["guard"])
+
+
+# Request/Response Models
+class ToolCallRequest(BaseModel):
+    """Tool call request."""
+    tool_name: str = Field(..., description="Name of the tool")
+    tool_input: Dict[str, Any] = Field(default_factory=dict, description="Tool parameters")
+    agent_id: str = Field(..., description="Agent making the request")
+    workspace_id: str = Field(..., description="Workspace context")
+    session_id: Optional[str] = Field(None, description="Session identifier")
+
+
+class ApprovalRequest(BaseModel):
+    """Request to approve a tool call."""
+    approval_id: str = Field(..., description="Approval request ID")
+    one_time: bool = Field(True, description="Whether this is a one-time approval")
+    expires_in_minutes: Optional[int] = Field(30, description="Approval expiration time")
+
+
+class DenyRequest(BaseModel):
+    """Request to deny a tool call."""
+    approval_id: str = Field(..., description="Approval request ID")
+    reason: Optional[str] = Field(None, description="Reason for denial")
+
+
+class ToolFinding(BaseModel):
+    """Tool guard finding."""
+    severity: SeverityLevel
+    message: str
+    field: Optional[str] = None
+
+
+class ApprovalResponse(BaseModel):
+    """Tool approval response."""
+    approval_id: str
+    status: ApprovalStatus
+    tool_name: str
+    tool_input: Dict[str, Any]
+    agent_id: str
+    workspace_id: str
+    session_id: Optional[str] = None
+    findings: List[ToolFinding] = Field(default_factory=list)
+    created_at: str
+    resolved_at: Optional[str] = None
+    resolved_by: Optional[str] = None
+
+
+class PendingApprovalsResponse(BaseModel):
+    """List of pending approvals."""
+    approvals: List[ApprovalResponse]
+    total: int
+
+
+STORE = TOOL_GUARD_STORE
+SAFE_TOOLS = {
+    "get_price",
+    "get_fundamentals",
+    "get_news",
+    "analyze_technical",
+}
+
+
+def _to_response(record: ApprovalRecord) -> ApprovalResponse:
+    return ApprovalResponse(
+        approval_id=record.approval_id,
+        status=record.status,
+        tool_name=record.tool_name,
+        tool_input=record.tool_input,
+        agent_id=record.agent_id,
+        workspace_id=record.workspace_id,
+        session_id=record.session_id,
+        findings=[ToolFinding(**f.to_dict()) for f in record.findings],
+        created_at=record.created_at.isoformat(),
+        resolved_at=record.resolved_at.isoformat() if record.resolved_at else None,
+        resolved_by=record.resolved_by,
+    )
+
+
+# Routes
+@router.post("/check", response_model=ApprovalResponse)
+async def check_tool_call(
+    request: ToolCallRequest,
+):
+    """
+    Check if a tool call requires approval.
+
+    Args:
+        request: Tool call details
+
+    Returns:
+        Approval status - may be auto-approved, auto-denied, or pending
+    """
+    record = STORE.create_pending(
+        tool_name=request.tool_name,
+        tool_input=request.tool_input,
+        agent_id=request.agent_id,
+        workspace_id=request.workspace_id,
+        session_id=request.session_id,
+        findings=default_findings_for_tool(request.tool_name),
+    )
+
+    if request.tool_name in SAFE_TOOLS:
+        record.status = ApprovalStatus.APPROVED
+        record.resolved_at = datetime.utcnow()
+        record.resolved_by = "system"
+        STORE.set_status(
+            record.approval_id,
+            ApprovalStatus.APPROVED,
+            resolved_by="system",
+            notify_request=False,
+        )
+
+    return _to_response(record)
+
+
+@router.post("/approve", response_model=ApprovalResponse)
+async def approve_tool_call(
+    request: ApprovalRequest,
+):
+    """
+    Approve a pending tool call.
+
+    Args:
+        request: Approval parameters
+
+    Returns:
+        Updated approval status
+    """
+    record = STORE.get(request.approval_id)
+    if not record:
+        raise HTTPException(status_code=404, detail="Approval request not found")
+
+    if record.status != ApprovalStatus.PENDING:
+        raise HTTPException(status_code=400, detail=f"Approval already {record.status}")
+
+    record.status = ApprovalStatus.APPROVED
+    record.resolved_at = datetime.utcnow()
+    record.resolved_by = "user"
+
+    return _to_response(record)
+
+
+@router.post("/deny", response_model=ApprovalResponse)
+async def deny_tool_call(
+    request: DenyRequest,
+):
+    """
+    Deny a pending tool call.
+
+    Args:
+        request: Denial parameters
+
+    Returns:
+        Updated approval status
+    """
+    record = STORE.get(request.approval_id)
+    if not record:
+        raise HTTPException(status_code=404, detail="Approval request not found")
+
+    if record.status != ApprovalStatus.PENDING:
+        raise HTTPException(status_code=400, detail=f"Approval already {record.status}")
+
+    record.status = ApprovalStatus.DENIED
+    record.resolved_at = datetime.utcnow()
+    record.resolved_by = "user"
+    record.metadata["denial_reason"] = request.reason
+
+    return _to_response(record)
+
+
+@router.get("/pending", response_model=PendingApprovalsResponse)
+async def list_pending_approvals(
+    workspace_id: Optional[str] = None,
+    agent_id: Optional[str] = None,
+):
+    """
+    List pending tool approval requests.
+
+    Args:
+        workspace_id: Filter by workspace
+        agent_id: Filter by agent
+
+    Returns:
+        List of pending approvals
+    """
+    pending = [
+        _to_response(record)
+        for record in STORE.list(
+            status=ApprovalStatus.PENDING,
+            workspace_id=workspace_id,
+            agent_id=agent_id,
+        )
+    ]
+    return PendingApprovalsResponse(approvals=pending, total=len(pending))
+
+
+@router.get("/approvals/{approval_id}", response_model=ApprovalResponse)
+async def get_approval_status(
+    approval_id: str,
+):
+    """
+    Get the status of a specific approval request.
+
+    Args:
+        approval_id: Approval request ID
+
+    Returns:
+        Approval status
+    """
+    record = STORE.get(approval_id)
+    if not record:
+        raise HTTPException(status_code=404, detail="Approval request not found")
+    return _to_response(record)
+
+
+@router.delete("/approvals/{approval_id}")
+async def cancel_approval(
+    approval_id: str,
+):
+    """
+    Cancel/delete a pending approval request.
+
+    Args:
+        approval_id: Approval request ID
+
+    Returns:
+        Success message
+    """
+    record = STORE.get(approval_id)
+    if not record:
+        raise HTTPException(status_code=404, detail="Approval request not found")
+
+    STORE.cancel(approval_id)
+    return _to_response(record)
--- a/backend/api/runtime.py
+++ b/backend/api/runtime.py
@@ -0,0 +1,711 @@
+# -*- coding: utf-8 -*-
+"""Runtime API routes - Control Plane for managing Gateway processes."""
+
+from __future__ import annotations
+
+import asyncio
+import json
+import logging
+import os
+import signal
+import subprocess
+import sys
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+logger = logging.getLogger(__name__)
+
+from fastapi import APIRouter, BackgroundTasks, HTTPException, Request
+from pydantic import BaseModel, Field
+
+from backend.runtime.agent_runtime import AgentRuntimeState
+from backend.config.bootstrap_config import (
+    resolve_runtime_config,
+    update_bootstrap_values_for_run,
+)
+
+router = APIRouter(prefix="/api/runtime", tags=["runtime"])
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+
+
+class RuntimeState:
+    """Thread-safe singleton for managing runtime state.
+
+    Encapsulates runtime_manager, _gateway_process, and _gateway_port
+    with asyncio.Lock protection for concurrent access.
+    """
+
+    _instance: Optional["RuntimeState"] = None
+    _lock: asyncio.Lock = asyncio.Lock()
+
+    def __new__(cls) -> "RuntimeState":
+        if cls._instance is None:
+            cls._instance = super().__new__(cls)
+            cls._instance._initialized = False
+        return cls._instance
+
+    def __init__(self) -> None:
+        if self._initialized:
+            return
+        self._runtime_manager: Optional[Any] = None
+        self._gateway_process: Optional[subprocess.Popen] = None
+        self._gateway_port: int = 8765
+        self._state_lock = asyncio.Lock()
+        self._initialized = True
+
+    @property
+    async def lock(self) -> asyncio.Lock:
+        """Get the asyncio lock for state synchronization."""
+        return self._state_lock
+
+    @property
+    def runtime_manager(self) -> Optional[Any]:
+        """Get the runtime manager (no lock - read only)."""
+        return self._runtime_manager
+
+    @runtime_manager.setter
+    def runtime_manager(self, value: Optional[Any]) -> None:
+        """Set the runtime manager."""
+        self._runtime_manager = value
+
+    @property
+    def gateway_process(self) -> Optional[subprocess.Popen]:
+        """Get the gateway process (no lock - read only)."""
+        return self._gateway_process
+
+    @gateway_process.setter
+    def gateway_process(self, value: Optional[subprocess.Popen]) -> None:
+        """Set the gateway process."""
+        self._gateway_process = value
+
+    @property
+    def gateway_port(self) -> int:
+        """Get the gateway port."""
+        return self._gateway_port
+
+    @gateway_port.setter
+    def gateway_port(self, value: int) -> None:
+        """Set the gateway port."""
+        self._gateway_port = value
+
+    async def set_runtime_manager(self, manager: Any) -> None:
+        """Set runtime manager with lock protection."""
+        async with self._state_lock:
+            self._runtime_manager = manager
+
+    async def get_runtime_manager(self) -> Optional[Any]:
+        """Get runtime manager with lock protection."""
+        async with self._state_lock:
+            return self._runtime_manager
+
+    async def set_gateway_process(self, process: Optional[subprocess.Popen]) -> None:
+        """Set gateway process with lock protection."""
+        async with self._state_lock:
+            self._gateway_process = process
+
+    async def get_gateway_process(self) -> Optional[subprocess.Popen]:
+        """Get gateway process with lock protection."""
+        async with self._state_lock:
+            return self._gateway_process
+
+    async def set_gateway_port(self, port: int) -> None:
+        """Set gateway port with lock protection."""
+        async with self._state_lock:
+            self._gateway_port = port
+
+    async def get_gateway_port(self) -> int:
+        """Get gateway port with lock protection."""
+        async with self._state_lock:
+            return self._gateway_port
+
+
+# Singleton instance
+_runtime_state = RuntimeState()
+
+
+def get_runtime_state() -> RuntimeState:
+    """Get the RuntimeState singleton instance."""
+    return _runtime_state
+
+
+# Backward compatibility: module-level runtime_manager for external imports
+# This is set by register_runtime_manager() for backward compatibility
+runtime_manager: Optional[Any] = None
+
+
+class RunContextResponse(BaseModel):
+    config_name: str
+    run_dir: str
+    bootstrap_values: Dict[str, Any]
+
+
+class RuntimeAgentState(BaseModel):
+    agent_id: str
+    status: str
+    last_session: Optional[str] = None
+    last_updated: str
+
+
+class RuntimeAgentsResponse(BaseModel):
+    agents: List[RuntimeAgentState]
+
+
+class RuntimeEvent(BaseModel):
+    timestamp: str
+    event: str
+    details: Dict[str, Any]
+    session: Optional[str]
+
+
+class RuntimeEventsResponse(BaseModel):
+    events: List[RuntimeEvent]
+
+
+class LaunchConfig(BaseModel):
+    """Configuration for launching a new trading task."""
+    tickers: List[str] = Field(default_factory=list, description="股票池")
+    schedule_mode: str = Field(default="daily", description="调度模式: daily, interval")
+    interval_minutes: int = Field(default=60, ge=1, description="间隔分钟数")
+    trigger_time: str = Field(default="09:30", description="触发时间 HH:MM")
+    max_comm_cycles: int = Field(default=2, ge=1, description="最大会商轮数")
+    initial_cash: float = Field(default=100000.0, gt=0, description="初始资金")
+    margin_requirement: float = Field(default=0.0, ge=0, description="保证金要求")
+    enable_memory: bool = Field(default=False, description="是否启用长期记忆")
+    mode: str = Field(default="live", description="运行模式: live, backtest")
+    start_date: Optional[str] = Field(default=None, description="回测开始日期 YYYY-MM-DD")
+    end_date: Optional[str] = Field(default=None, description="回测结束日期 YYYY-MM-DD")
+    poll_interval: int = Field(default=10, ge=1, le=300, description="市场数据轮询间隔(秒)")
+    enable_mock: bool = Field(default=False, description="是否启用模拟模式(使用模拟价格数据)")
+
+
+class LaunchResponse(BaseModel):
+    run_id: str
+    status: str
+    run_dir: str
+    gateway_port: int
+    message: str
+
+
+class StopResponse(BaseModel):
+    status: str
+    message: str
+
+
+class GatewayStatusResponse(BaseModel):
+    is_running: bool
+    port: int
+    run_id: Optional[str] = None
+
+
+class RuntimeConfigResponse(BaseModel):
+    run_id: str
+    is_running: bool
+    gateway_port: int
+    bootstrap: Dict[str, Any]
+    resolved: Dict[str, Any]
+
+
+class UpdateRuntimeConfigRequest(BaseModel):
+    schedule_mode: Optional[str] = None
+    interval_minutes: Optional[int] = Field(default=None, ge=1)
+    trigger_time: Optional[str] = None
+    max_comm_cycles: Optional[int] = Field(default=None, ge=1)
+    initial_cash: Optional[float] = Field(default=None, gt=0)
+    margin_requirement: Optional[float] = Field(default=None, ge=0)
+    enable_memory: Optional[bool] = None
+
+
+def _generate_run_id() -> str:
+    """Generate timestamp-based run ID: YYYYMMDD_HHMMSS"""
+    return datetime.now().strftime("%Y%m%d_%H%M%S")
+
+
+def _get_run_dir(run_id: str) -> Path:
+    """Return the run directory for a given run ID."""
+    return PROJECT_ROOT / "runs" / run_id
+
+
+def _find_available_port(start_port: int = 8765, max_port: int = 9000) -> int:
+    """Find an available port for Gateway."""
+    import socket
+    for port in range(start_port, max_port):
+        with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s:
+            if s.connect_ex(('localhost', port)) != 0:
+                return port
+    raise RuntimeError("No available port found")
+
+
+def _is_gateway_running() -> bool:
+    """Check if Gateway process is running."""
+    process = _runtime_state.gateway_process
+    if process is None:
+        return False
+    return process.poll() is None
+
+
+def _stop_gateway() -> bool:
+    """Stop the Gateway process."""
+    process = _runtime_state.gateway_process
+    if process is None:
+        return False
+
+    try:
+        # Try graceful shutdown first
+        process.terminate()
+        try:
+            process.wait(timeout=5)
+        except subprocess.TimeoutExpired:
+            # Force kill if graceful shutdown fails
+            process.kill()
+            process.wait()
+    except Exception as e:
+        logger.warning(f"Error during gateway shutdown: {e}")
+    finally:
+        _runtime_state.gateway_process = None
+
+    return True
+
+
+def _start_gateway_process(
+    run_id: str,
+    run_dir: Path,
+    bootstrap: Dict[str, Any],
+    port: int
+) -> subprocess.Popen:
+    """Start Gateway as a separate process."""
+    # Prepare environment
+    env = os.environ.copy()
+
+    # Create command arguments
+    cmd = [
+        sys.executable,
+        "-m", "backend.gateway_server",
+        "--run-id", run_id,
+        "--run-dir", str(run_dir),
+        "--port", str(port),
+        "--bootstrap", json.dumps(bootstrap)
+    ]
+
+    # Start process
+    process = subprocess.Popen(
+        cmd,
+        env=env,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        cwd=PROJECT_ROOT
+    )
+
+    return process
+
+
+@router.get("/context", response_model=RunContextResponse)
+async def get_run_context() -> RunContextResponse:
+    """Return the current run context from in-memory state (avoids glob race condition)."""
+    manager = _runtime_state.runtime_manager
+    if manager is None or manager.context is None:
+        raise HTTPException(status_code=404, detail="No run context available")
+
+    context = manager.context
+    return RunContextResponse(
+        config_name=context.config_name,
+        run_dir=str(context.run_dir),
+        bootstrap_values=context.bootstrap_values,
+    )
+
+
+@router.get("/agents", response_model=RuntimeAgentsResponse)
+async def get_runtime_agents() -> RuntimeAgentsResponse:
+    """Return agent states from the in-memory runtime manager (avoids glob race condition)."""
+    manager = _runtime_state.runtime_manager
+    if manager is None:
+        raise HTTPException(status_code=404, detail="No runtime state available")
+
+    snapshot = manager.build_snapshot()
+    agents = snapshot.get("agents", [])
+
+    return RuntimeAgentsResponse(
+        agents=[RuntimeAgentState(**a) for a in agents]
+    )
+
+
+@router.get("/events", response_model=RuntimeEventsResponse)
+async def get_runtime_events() -> RuntimeEventsResponse:
+    """Return events from the in-memory runtime manager (avoids glob race condition)."""
+    manager = _runtime_state.runtime_manager
+    if manager is None:
+        raise HTTPException(status_code=404, detail="No runtime state available")
+
+    snapshot = manager.build_snapshot()
+    events = snapshot.get("events", [])
+
+    return RuntimeEventsResponse(
+        events=[RuntimeEvent(**e) for e in events]
+    )
+
+
+@router.get("/gateway/status", response_model=GatewayStatusResponse)
+async def get_gateway_status() -> GatewayStatusResponse:
+    """Get Gateway process status and port."""
+    is_running = _is_gateway_running()
+    run_id = None
+
+    if is_running:
+        # Get run_id from in-memory runtime manager (avoids glob race condition)
+        manager = _runtime_state.runtime_manager
+        if manager is not None and manager.context is not None:
+            run_id = manager.context.config_name
+
+    return GatewayStatusResponse(
+        is_running=is_running,
+        port=_runtime_state.gateway_port,
+        run_id=run_id
+    )
+
+
+@router.get("/gateway/port")
+async def get_gateway_port(request: Request) -> Dict[str, Any]:
+    """Get WebSocket Gateway port for frontend connection."""
+    gateway_port = _runtime_state.gateway_port
+    return {
+        "port": gateway_port,
+        "is_running": _is_gateway_running(),
+        "ws_url": _build_gateway_ws_url(request, gateway_port),
+    }
+
+
+def _build_gateway_ws_url(request: Request, port: int) -> str:
+    """Build a proxy-safe Gateway WebSocket URL."""
+    forwarded_proto = request.headers.get("x-forwarded-proto", "").split(",")[0].strip()
+    scheme = forwarded_proto or request.url.scheme
+    ws_scheme = "wss" if scheme == "https" else "ws"
+
+    forwarded_host = request.headers.get("x-forwarded-host", "").split(",")[0].strip()
+    host = forwarded_host or request.url.hostname or "localhost"
+    if ":" in host and not host.startswith("["):
+        host = host.split(":", 1)[0]
+
+    return f"{ws_scheme}://{host}:{port}"
+
+
+def _get_current_runtime_context() -> Dict[str, Any]:
+    """Return the active runtime context from the in-memory manager (avoids glob race condition).
+
+    Falls back to file-based lookup only when the in-memory manager is not available
+    (e.g., after a service restart). File-based lookup is deprecated and exists
+    only for backward compatibility.
+    """
+    if not _is_gateway_running():
+        raise HTTPException(status_code=404, detail="No runtime is currently running")
+
+    # Primary: use in-memory manager (always correct for current process)
+    manager = _runtime_state.runtime_manager
+    if manager is not None and manager.context is not None:
+        ctx = manager.context
+        return {
+            "config_name": ctx.config_name,
+            "run_dir": str(ctx.run_dir),
+            "bootstrap_values": ctx.bootstrap_values,
+        }
+
+    # Deprecated fallback: scan filesystem (only for backward compatibility
+    # after service restart without a restart of the runtime itself)
+    snapshots = sorted(
+        PROJECT_ROOT.glob("runs/*/state/runtime_state.json"),
+        key=lambda p: p.stat().st_mtime,
+        reverse=True,
+    )
+    if not snapshots:
+        raise HTTPException(status_code=404, detail="No runtime information available")
+    latest = json.loads(snapshots[0].read_text(encoding="utf-8"))
+    context = latest.get("context") or {}
+    if not context.get("config_name"):
+        raise HTTPException(status_code=404, detail="No runtime context available")
+    return context
+
+
+def _resolve_runtime_response(run_id: str) -> RuntimeConfigResponse:
+    """Build a normalized runtime config response for the active run."""
+    context = _get_current_runtime_context()
+    bootstrap = dict(context.get("bootstrap_values") or {})
+    resolved = resolve_runtime_config(
+        project_root=PROJECT_ROOT,
+        config_name=run_id,
+        enable_memory=bool(bootstrap.get("enable_memory", False)),
+        schedule_mode=str(bootstrap.get("schedule_mode", "daily")),
+        interval_minutes=int(bootstrap.get("interval_minutes", 60) or 60),
+        trigger_time=str(bootstrap.get("trigger_time", "09:30") or "09:30"),
+    )
+    return RuntimeConfigResponse(
+        run_id=run_id,
+        is_running=True,
+        gateway_port=_runtime_state.gateway_port,
+        bootstrap=bootstrap,
+        resolved=resolved,
+    )
+
+
+def _normalize_runtime_config_updates(
+    request: UpdateRuntimeConfigRequest,
+) -> Dict[str, Any]:
+    """Validate and normalize runtime config updates."""
+    updates: Dict[str, Any] = {}
+
+    if request.schedule_mode is not None:
+        schedule_mode = str(request.schedule_mode).strip().lower()
+        if schedule_mode not in {"daily", "intraday"}:
+            raise HTTPException(
+                status_code=400,
+                detail="schedule_mode must be 'daily' or 'intraday'",
+            )
+        updates["schedule_mode"] = schedule_mode
+
+    if request.interval_minutes is not None:
+        updates["interval_minutes"] = int(request.interval_minutes)
+
+    if request.trigger_time is not None:
+        trigger_time = str(request.trigger_time).strip()
+        if trigger_time and trigger_time != "now":
+            try:
+                datetime.strptime(trigger_time, "%H:%M")
+            except ValueError as exc:
+                raise HTTPException(
+                    status_code=400,
+                    detail="trigger_time must use HH:MM or 'now'",
+                ) from exc
+        updates["trigger_time"] = trigger_time or "09:30"
+
+    if request.max_comm_cycles is not None:
+        updates["max_comm_cycles"] = int(request.max_comm_cycles)
+
+    if request.initial_cash is not None:
+        updates["initial_cash"] = float(request.initial_cash)
+
+    if request.margin_requirement is not None:
+        updates["margin_requirement"] = float(request.margin_requirement)
+
+    if request.enable_memory is not None:
+        updates["enable_memory"] = bool(request.enable_memory)
+
+    if not updates:
+        raise HTTPException(status_code=400, detail="No runtime config updates provided")
+
+    return updates
+
+
+@router.post("/start", response_model=LaunchResponse)
+async def start_runtime(
+    config: LaunchConfig,
+    background_tasks: BackgroundTasks
+) -> LaunchResponse:
+    """Start a new trading runtime with the given configuration.
+
+    1. Stop existing Gateway if running
+    2. Generate run ID and directory
+    3. Create runtime manager
+    4. Start Gateway as subprocess (Data Plane)
+    5. Return Gateway port for WebSocket connection
+    """
+    # Lazy import to avoid circular dependency
+    from backend.runtime.manager import TradingRuntimeManager
+
+    # 1. Stop existing Gateway
+    if _is_gateway_running():
+        _stop_gateway()
+        await asyncio.sleep(1)  # Wait for port release
+
+    # 2. Generate run ID and directory
+    run_id = _generate_run_id()
+    run_dir = _get_run_dir(run_id)
+
+    # 3. Prepare bootstrap config
+    bootstrap = {
+        "tickers": config.tickers,
+        "schedule_mode": config.schedule_mode,
+        "interval_minutes": config.interval_minutes,
+        "trigger_time": config.trigger_time,
+        "max_comm_cycles": config.max_comm_cycles,
+        "initial_cash": config.initial_cash,
+        "margin_requirement": config.margin_requirement,
+        "enable_memory": config.enable_memory,
+        "mode": config.mode,
+        "start_date": config.start_date,
+        "end_date": config.end_date,
+        "poll_interval": config.poll_interval,
+        "enable_mock": config.enable_mock,
+    }
+
+    # 4. Create runtime manager
+    manager = TradingRuntimeManager(
+        config_name=run_id,
+        run_dir=run_dir,
+        bootstrap=bootstrap,
+    )
+    manager.prepare_run()
+    register_runtime_manager(manager)
+
+    # 5. Write BOOTSTRAP.md
+    _write_bootstrap_md(run_dir, bootstrap)
+
+    # 6. Find available port and start Gateway process
+    gateway_port = _find_available_port(start_port=8765)
+    _runtime_state.gateway_port = gateway_port
+
+    try:
+        process = _start_gateway_process(
+            run_id=run_id,
+            run_dir=run_dir,
+            bootstrap=bootstrap,
+            port=gateway_port
+        )
+        _runtime_state.gateway_process = process
+
+        # Wait briefly to check if process started successfully
+        await asyncio.sleep(2)
+
+        if not _is_gateway_running():
+            stdout, stderr = process.communicate(timeout=1)
+            _runtime_state.gateway_process = None
+            raise HTTPException(
+                status_code=500,
+                detail=f"Gateway failed to start: {stderr.decode() if stderr else 'Unknown error'}"
+            )
+
+    except Exception as e:
+        _stop_gateway()
+        raise HTTPException(status_code=500, detail=f"Failed to start Gateway: {str(e)}")
+
+    return LaunchResponse(
+        run_id=run_id,
+        status="started",
+        run_dir=str(run_dir),
+        gateway_port=gateway_port,
+        message=f"Runtime started with run_id: {run_id}, Gateway on port: {gateway_port}",
+    )
+
+
+@router.get("/config", response_model=RuntimeConfigResponse)
+async def get_runtime_config() -> RuntimeConfigResponse:
+    """Return the current runtime bootstrap and resolved settings."""
+    context = _get_current_runtime_context()
+    return _resolve_runtime_response(context["config_name"])
+
+
+@router.put("/config", response_model=RuntimeConfigResponse)
+async def update_runtime_config(
+    request: UpdateRuntimeConfigRequest,
+) -> RuntimeConfigResponse:
+    """Persist selected runtime configuration updates for the active run."""
+    context = _get_current_runtime_context()
+    run_id = context["config_name"]
+    updates = _normalize_runtime_config_updates(request)
+    updated = update_bootstrap_values_for_run(PROJECT_ROOT, run_id, updates)
+
+    manager = _runtime_state.runtime_manager
+    if manager is not None and getattr(manager, "config_name", None) == run_id:
+        manager.bootstrap.update(updates)
+        if getattr(manager, "context", None) is not None:
+            manager.context.bootstrap_values.update(updates)
+        if hasattr(manager, "_persist_snapshot"):
+            manager._persist_snapshot()
+
+    response = _resolve_runtime_response(run_id)
+    response.bootstrap = dict(updated.values)
+    return response
+
+
+@router.post("/stop", response_model=StopResponse)
+async def stop_runtime(force: bool = True) -> StopResponse:
+    """Stop the current running runtime."""
+    was_running = _is_gateway_running()
+
+    if not was_running:
+        raise HTTPException(status_code=404, detail="No runtime is currently running")
+
+    # Stop Gateway process
+    _stop_gateway()
+
+    # Unregister runtime manager
+    unregister_runtime_manager()
+
+    return StopResponse(
+        status="stopped",
+        message="Runtime stopped successfully",
+    )
+
+
+@router.post("/restart")
+async def restart_runtime(
+    config: LaunchConfig,
+    background_tasks: BackgroundTasks
+):
+    """Restart the runtime with a new configuration."""
+    # Stop current runtime
+    await stop_runtime(force=True)
+
+    # Start new runtime
+    response = await start_runtime(config, background_tasks)
+
+    return {
+        "run_id": response.run_id,
+        "status": "restarted",
+        "gateway_port": response.gateway_port,
+        "message": f"Runtime restarted with run_id: {response.run_id}",
+    }
+
+
+@router.get("/current")
+async def get_current_runtime():
+    """Get information about the currently running runtime."""
+    if not _is_gateway_running():
+        raise HTTPException(status_code=404, detail="No runtime is currently running")
+
+    # Get context from in-memory manager (avoids glob race condition)
+    context = _get_current_runtime_context()
+
+    return {
+        "run_id": context.get("config_name"),
+        "run_dir": context.get("run_dir"),
+        "is_running": True,
+        "gateway_port": _runtime_state.gateway_port,
+        "bootstrap": context.get("bootstrap_values", {}),
+    }
+
+
+def register_runtime_manager(manager: Any) -> None:
+    """Allow other modules to expose the runtime manager to the API."""
+    global runtime_manager
+    runtime_manager = manager
+    # Also update the RuntimeState singleton for internal consistency
+    _runtime_state.runtime_manager = manager
+
+
+def unregister_runtime_manager() -> None:
+    """Drop the runtime manager reference."""
+    global runtime_manager
+    runtime_manager = None
+    # Also update the RuntimeState singleton for internal consistency
+    _runtime_state.runtime_manager = None
+
+
+def _write_bootstrap_md(run_dir: Path, bootstrap: Dict[str, Any]) -> None:
+    """Write bootstrap configuration to BOOTSTRAP.md."""
+    try:
+        import yaml
+    except ImportError:
+        yaml = None
+
+    bootstrap_path = run_dir / "BOOTSTRAP.md"
+    bootstrap_path.parent.mkdir(parents=True, exist_ok=True)
+
+    # Filter out None values
+    values = {k: v for k, v in bootstrap.items() if v is not None}
+
+    if yaml:
+        front_matter = yaml.safe_dump(values, allow_unicode=True, sort_keys=False)
+    else:
+        front_matter = json.dumps(values, ensure_ascii=False, indent=2)
+
+    content = f"---\n{front_matter}---\n"
+    bootstrap_path.write_text(content, encoding="utf-8")
--- a/backend/api/workspaces.py
+++ b/backend/api/workspaces.py
@@ -0,0 +1,196 @@
+# -*- coding: utf-8 -*-
+"""
+Workspace API Routes
+
+Provides REST API endpoints for workspace management.
+"""
+from typing import Any, Dict, List, Optional
+
+from fastapi import APIRouter, HTTPException, Depends
+from pydantic import BaseModel, Field
+
+from backend.agents import WorkspaceManager
+
+router = APIRouter(prefix="/api/workspaces", tags=["workspaces"])
+
+
+# Request/Response Models
+class CreateWorkspaceRequest(BaseModel):
+    """Request to create a new workspace."""
+    workspace_id: str = Field(..., description="Unique workspace identifier")
+    name: Optional[str] = Field(None, description="Display name")
+    description: Optional[str] = Field(None, description="Workspace description")
+    metadata: Optional[Dict[str, Any]] = Field(None, description="Additional metadata")
+
+
+class UpdateWorkspaceRequest(BaseModel):
+    """Request to update a workspace."""
+    name: Optional[str] = None
+    description: Optional[str] = None
+    metadata: Optional[Dict[str, Any]] = None
+
+
+class WorkspaceResponse(BaseModel):
+    """Workspace information response."""
+    workspace_id: str
+    name: str
+    description: str
+    created_at: Optional[str] = None
+    metadata: Dict[str, Any] = Field(default_factory=dict)
+
+
+class WorkspaceListResponse(BaseModel):
+    """List of workspaces response."""
+    workspaces: List[WorkspaceResponse]
+    total: int
+
+
+# Dependencies
+def get_workspace_manager():
+    """Get WorkspaceManager instance."""
+    return WorkspaceManager()
+
+
+# Routes
+@router.post("", response_model=WorkspaceResponse)
+async def create_workspace(
+    request: CreateWorkspaceRequest,
+    manager: WorkspaceManager = Depends(get_workspace_manager),
+):
+    """
+    Create a new workspace.
+
+    Args:
+        request: Workspace creation parameters
+
+    Returns:
+        Created workspace information
+    """
+    try:
+        config = manager.create_workspace(
+            workspace_id=request.workspace_id,
+            name=request.name,
+            description=request.description,
+            metadata=request.metadata or {},
+        )
+        return WorkspaceResponse(
+            workspace_id=config.workspace_id,
+            name=config.name,
+            description=config.description,
+            created_at=config.created_at,
+            metadata=config.metadata,
+        )
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
+
+
+@router.get("", response_model=WorkspaceListResponse)
+async def list_workspaces(
+    manager: WorkspaceManager = Depends(get_workspace_manager),
+):
+    """
+    List all workspaces.
+
+    Returns:
+        List of workspaces
+    """
+    workspaces = manager.list_workspaces()
+    return WorkspaceListResponse(
+        workspaces=[
+            WorkspaceResponse(
+                workspace_id=ws.workspace_id,
+                name=ws.name,
+                description=ws.description,
+                created_at=ws.created_at,
+                metadata=ws.metadata,
+            )
+            for ws in workspaces
+        ],
+        total=len(workspaces),
+    )
+
+
+@router.get("/{workspace_id}", response_model=WorkspaceResponse)
+async def get_workspace(
+    workspace_id: str,
+    manager: WorkspaceManager = Depends(get_workspace_manager),
+):
+    """
+    Get workspace details.
+
+    Args:
+        workspace_id: Workspace identifier
+
+    Returns:
+        Workspace information
+    """
+    workspace = manager.get_workspace(workspace_id)
+    if not workspace:
+        raise HTTPException(status_code=404, detail=f"Workspace '{workspace_id}' not found")
+
+    return WorkspaceResponse(
+        workspace_id=workspace["workspace_id"],
+        name=workspace.get("name", workspace_id),
+        description=workspace.get("description", ""),
+        created_at=workspace.get("created_at"),
+        metadata=workspace.get("metadata", {}),
+    )
+
+
+@router.patch("/{workspace_id}", response_model=WorkspaceResponse)
+async def update_workspace(
+    workspace_id: str,
+    request: UpdateWorkspaceRequest,
+    manager: WorkspaceManager = Depends(get_workspace_manager),
+):
+    """
+    Update workspace configuration.
+
+    Args:
+        workspace_id: Workspace identifier
+        request: Update parameters
+
+    Returns:
+        Updated workspace information
+    """
+    try:
+        config = manager.update_workspace_config(
+            workspace_id=workspace_id,
+            name=request.name,
+            description=request.description,
+            metadata=request.metadata,
+        )
+        return WorkspaceResponse(
+            workspace_id=config.workspace_id,
+            name=config.name,
+            description=config.description,
+            created_at=config.created_at,
+            metadata=config.metadata,
+        )
+    except ValueError as e:
+        raise HTTPException(status_code=404, detail=str(e))
+
+
+@router.delete("/{workspace_id}")
+async def delete_workspace(
+    workspace_id: str,
+    force: bool = False,
+    manager: WorkspaceManager = Depends(get_workspace_manager),
+):
+    """
+    Delete a workspace.
+
+    Args:
+        workspace_id: Workspace identifier
+        force: If True, delete even if workspace has agents
+
+    Returns:
+        Success message
+    """
+    try:
+        success = manager.delete_workspace(workspace_id, force=force)
+        if not success:
+            raise HTTPException(status_code=404, detail=f"Workspace '{workspace_id}' not found")
+        return {"message": f"Workspace '{workspace_id}' deleted successfully"}
+    except ValueError as e:
+        raise HTTPException(status_code=400, detail=str(e))
--- a/backend/apps/init.py
+++ b/backend/apps/init.py
@@ -0,0 +1,27 @@
+# -*- coding: utf-8 -*-
+"""Application surfaces for progressive service extraction."""
+
+from .agent_service import app as agent_app
+from .agent_service import create_app as create_agent_app
+from .news_service import app as news_app
+from .news_service import create_app as create_news_app
+from .runtime_service import app as runtime_app
+from .runtime_service import create_app as create_runtime_app
+from .trading_service import app as trading_app
+from .trading_service import create_app as create_trading_app
+
+app = agent_app
+create_app = create_agent_app
+
+__all__ = [
+    "app",
+    "create_app",
+    "agent_app",
+    "create_agent_app",
+    "news_app",
+    "create_news_app",
+    "runtime_app",
+    "create_runtime_app",
+    "trading_app",
+    "create_trading_app",
+]
--- a/backend/apps/agent_service.py
+++ b/backend/apps/agent_service.py
@@ -0,0 +1,95 @@
+# -*- coding: utf-8 -*-
+"""Agent control-plane FastAPI surface."""
+
+from __future__ import annotations
+
+from contextlib import asynccontextmanager
+from pathlib import Path
+from typing import AsyncGenerator
+
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+
+from backend.api import agents_router, guard_router, workspaces_router
+from backend.agents import AgentFactory, WorkspaceManager, get_registry
+from backend.config.env_config import get_cors_origins
+
+# Global instances (initialized on startup)
+agent_factory: AgentFactory | None = None
+workspace_manager: WorkspaceManager | None = None
+
+
+def create_app(project_root: Path | None = None) -> FastAPI:
+    """Create the agent control-plane app."""
+    resolved_project_root = project_root or Path(__file__).resolve().parents[2]
+
+    @asynccontextmanager
+    async def lifespan(_app: FastAPI) -> AsyncGenerator[None, None]:
+        """Initialize workspace and registry state for the control plane."""
+        global agent_factory, workspace_manager
+
+        workspace_manager = WorkspaceManager(project_root=resolved_project_root)
+        agent_factory = AgentFactory(project_root=resolved_project_root)
+        agent_factory.workspaces_root.mkdir(parents=True, exist_ok=True)
+
+        registry = get_registry()
+        print("✓ EvoTraders API started")
+        print(f"  - Workspaces root: {agent_factory.workspaces_root}")
+        print(f"  - Registered agents: {registry.get_agent_count()}")
+
+        yield
+
+        print("✓ EvoTraders API shutting down")
+
+    app = FastAPI(
+        title="EvoTraders Agent Service",
+        description="REST API for the EvoTraders multi-agent control plane",
+        version="0.1.0",
+        lifespan=lifespan,
+    )
+
+    app.add_middleware(
+        CORSMiddleware,
+        allow_origins=get_cors_origins(),
+        allow_credentials=True,
+        allow_methods=["*"],
+        allow_headers=["*"],
+    )
+
+    @app.get("/health")
+    async def health_check() -> dict[str, object]:
+        """Health check endpoint."""
+        registry = get_registry()
+        return {
+            "status": "healthy",
+            "version": "0.1.0",
+            "agents_registered": registry.get_agent_count(),
+            "workspaces_available": (
+                len(workspace_manager.list_workspaces())
+                if workspace_manager
+                else 0
+            ),
+        }
+
+    @app.get("/api/status")
+    async def api_status() -> dict[str, object]:
+        """Get API status and registry information."""
+        registry = get_registry()
+        return {
+            "status": "operational",
+            "registry": registry.get_stats(),
+        }
+
+    app.include_router(workspaces_router)
+    app.include_router(agents_router)
+    app.include_router(guard_router)
+    return app
+
+
+app = create_app()
+
+
+if __name__ == "__main__":
+    import uvicorn
+
+    uvicorn.run(app, host="0.0.0.0", port=8000)
--- a/backend/apps/news_service.py
+++ b/backend/apps/news_service.py
@@ -0,0 +1,154 @@
+# -*- coding: utf-8 -*-
+"""News and explain FastAPI surface."""
+
+from __future__ import annotations
+
+from typing import Any
+
+from fastapi import Depends, FastAPI, Query
+from fastapi.middleware.cors import CORSMiddleware
+
+from backend.data.market_store import MarketStore
+from backend.domains import news as news_domain
+from backend.config.env_config import get_cors_origins
+
+
+def get_market_store() -> MarketStore:
+    """Create a market store dependency."""
+    return MarketStore()
+
+
+def create_app() -> FastAPI:
+    """Create the news/explain service app."""
+    app = FastAPI(
+        title="EvoTraders News Service",
+        description="Read-only news enrichment and explain service surface extracted from the monolith",
+        version="0.1.0",
+    )
+
+    app.add_middleware(
+        CORSMiddleware,
+        allow_origins=get_cors_origins(),
+        allow_credentials=True,
+        allow_methods=["*"],
+        allow_headers=["*"],
+    )
+
+    @app.get("/health")
+    async def health_check() -> dict[str, str]:
+        return {"status": "healthy", "service": "news-service"}
+
+    @app.get("/api/enriched-news")
+    async def api_get_enriched_news(
+        ticker: str = Query(..., min_length=1),
+        start_date: str | None = Query(None),
+        end_date: str | None = Query(None),
+        limit: int = Query(100, ge=1, le=1000),
+        store: MarketStore = Depends(get_market_store),
+    ) -> dict[str, Any]:
+        return news_domain.get_enriched_news(
+            store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        )
+
+    @app.get("/api/news-for-date")
+    async def api_get_news_for_date(
+        ticker: str = Query(..., min_length=1),
+        date: str = Query(...),
+        limit: int = Query(20, ge=1, le=100),
+        store: MarketStore = Depends(get_market_store),
+    ) -> dict[str, Any]:
+        return news_domain.get_news_for_date(
+            store,
+            ticker=ticker,
+            date=date,
+            limit=limit,
+        )
+
+    @app.get("/api/news-timeline")
+    async def api_get_news_timeline(
+        ticker: str = Query(..., min_length=1),
+        start_date: str = Query(...),
+        end_date: str = Query(...),
+        store: MarketStore = Depends(get_market_store),
+    ) -> dict[str, Any]:
+        return news_domain.get_news_timeline(
+            store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+        )
+
+    @app.get("/api/categories")
+    async def api_get_categories(
+        ticker: str = Query(..., min_length=1),
+        start_date: str | None = Query(None),
+        end_date: str | None = Query(None),
+        limit: int = Query(200, ge=1, le=1000),
+        store: MarketStore = Depends(get_market_store),
+    ) -> dict[str, Any]:
+        return news_domain.get_news_categories(
+            store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        )
+
+    @app.get("/api/similar-days")
+    async def api_get_similar_days(
+        ticker: str = Query(..., min_length=1),
+        date: str = Query(...),
+        n_similar: int = Query(5, ge=1, le=20),
+        store: MarketStore = Depends(get_market_store),
+    ) -> dict[str, Any]:
+        return news_domain.get_similar_days_payload(
+            store,
+            ticker=ticker,
+            date=date,
+            n_similar=n_similar,
+        )
+
+    @app.get("/api/stories/{ticker}")
+    async def api_get_story(
+        ticker: str,
+        as_of_date: str = Query(...),
+        store: MarketStore = Depends(get_market_store),
+    ) -> dict[str, Any]:
+        return news_domain.get_story_payload(
+            store,
+            ticker=ticker,
+            as_of_date=as_of_date,
+        )
+
+    @app.get("/api/range-explain")
+    async def api_get_range_explain(
+        ticker: str = Query(..., min_length=1),
+        start_date: str = Query(...),
+        end_date: str = Query(...),
+        article_ids: list[str] = Query(default=[]),
+        limit: int = Query(100, ge=1, le=500),
+        store: MarketStore = Depends(get_market_store),
+    ) -> dict[str, Any]:
+        return news_domain.get_range_explain_payload(
+            store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            article_ids=article_ids,
+            limit=limit,
+        )
+
+    return app
+
+
+app = create_app()
+
+
+if __name__ == "__main__":
+    import uvicorn
+
+    uvicorn.run(app, host="0.0.0.0", port=8002)
--- a/backend/apps/runtime_service.py
+++ b/backend/apps/runtime_service.py
@@ -0,0 +1,69 @@
+# -*- coding: utf-8 -*-
+"""Dedicated runtime service FastAPI surface."""
+
+from __future__ import annotations
+
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+
+from backend.api import runtime_router
+from backend.api.runtime import get_runtime_state
+from backend.config.env_config import get_cors_origins
+
+
+def create_app() -> FastAPI:
+    """Create the runtime service app."""
+    app = FastAPI(
+        title="EvoTraders Runtime Service",
+        description="Runtime lifecycle and gateway service surface extracted from the monolith",
+        version="0.1.0",
+    )
+
+    app.add_middleware(
+        CORSMiddleware,
+        allow_origins=get_cors_origins(),
+        allow_credentials=True,
+        allow_methods=["*"],
+        allow_headers=["*"],
+    )
+
+    @app.get("/health")
+    async def health_check() -> dict[str, object]:
+        """Health check for the runtime service."""
+        runtime_state = get_runtime_state()
+        process = runtime_state.gateway_process
+        is_running = process is not None and process.poll() is None
+        return {
+            "status": "healthy",
+            "service": "runtime-service",
+            "gateway_running": is_running,
+            "gateway_port": runtime_state.gateway_port,
+        }
+
+    @app.get("/api/status")
+    async def api_status() -> dict[str, object]:
+        """Service-level status payload for runtime orchestration."""
+        runtime_state = get_runtime_state()
+        process = runtime_state.gateway_process
+        is_running = process is not None and process.poll() is None
+        return {
+            "status": "operational",
+            "service": "runtime-service",
+            "runtime": {
+                "gateway_running": is_running,
+                "gateway_port": runtime_state.gateway_port,
+                "has_runtime_manager": runtime_state.runtime_manager is not None,
+            },
+        }
+
+    app.include_router(runtime_router)
+    return app
+
+
+app = create_app()
+
+
+if __name__ == "__main__":
+    import uvicorn
+
+    uvicorn.run(app, host="0.0.0.0", port=8003)
--- a/backend/apps/trading_service.py
+++ b/backend/apps/trading_service.py
@@ -0,0 +1,150 @@
+# -*- coding: utf-8 -*-
+"""Trading data FastAPI surface."""
+
+from __future__ import annotations
+
+from typing import Any
+
+from fastapi import FastAPI, Query
+from fastapi.middleware.cors import CORSMiddleware
+
+from backend.config.env_config import get_cors_origins
+from backend.services.market import MarketService
+from backend.tools.data_tools import (
+    get_company_news,
+    get_financial_metrics,
+    get_insider_trades,
+    get_market_cap,
+    get_prices,
+    search_line_items,
+)
+from shared.schema import (
+    CompanyNewsResponse,
+    FinancialMetricsResponse,
+    InsiderTradeResponse,
+    LineItemResponse,
+    PriceResponse,
+)
+
+
+def create_app() -> FastAPI:
+    """Create the trading data service app."""
+    app = FastAPI(
+        title="EvoTraders Trading Service",
+        description="Read-only trading data service surface extracted from the monolith",
+        version="0.1.0",
+    )
+
+    app.add_middleware(
+        CORSMiddleware,
+        allow_origins=get_cors_origins(),
+        allow_credentials=True,
+        allow_methods=["*"],
+        allow_headers=["*"],
+    )
+
+    @app.get("/health")
+    async def health_check() -> dict[str, str]:
+        """Health check endpoint."""
+        return {"status": "healthy", "service": "trading-service"}
+
+    @app.get("/api/prices", response_model=PriceResponse)
+    async def api_get_prices(
+        ticker: str = Query(..., min_length=1),
+        start_date: str = Query(...),
+        end_date: str = Query(...),
+    ) -> PriceResponse:
+        prices = get_prices(ticker=ticker, start_date=start_date, end_date=end_date)
+        return PriceResponse(ticker=ticker, prices=prices)
+
+    @app.get("/api/financials", response_model=FinancialMetricsResponse)
+    async def api_get_financials(
+        ticker: str = Query(..., min_length=1),
+        end_date: str = Query(...),
+        period: str = Query("ttm"),
+        limit: int = Query(10, ge=1, le=100),
+    ) -> FinancialMetricsResponse:
+        metrics = get_financial_metrics(
+            ticker=ticker,
+            end_date=end_date,
+            period=period,
+            limit=limit,
+        )
+        return FinancialMetricsResponse(financial_metrics=metrics)
+
+    @app.get("/api/news", response_model=CompanyNewsResponse)
+    async def api_get_news(
+        ticker: str = Query(..., min_length=1),
+        end_date: str = Query(...),
+        start_date: str | None = Query(None),
+        limit: int = Query(1000, ge=1, le=5000),
+    ) -> CompanyNewsResponse:
+        news = get_company_news(
+            ticker=ticker,
+            end_date=end_date,
+            start_date=start_date,
+            limit=limit,
+        )
+        return CompanyNewsResponse(news=news)
+
+    @app.get("/api/insider-trades", response_model=InsiderTradeResponse)
+    async def api_get_insider_trades(
+        ticker: str = Query(..., min_length=1),
+        end_date: str = Query(...),
+        start_date: str | None = Query(None),
+        limit: int = Query(1000, ge=1, le=5000),
+    ) -> InsiderTradeResponse:
+        trades = get_insider_trades(
+            ticker=ticker,
+            end_date=end_date,
+            start_date=start_date,
+            limit=limit,
+        )
+        return InsiderTradeResponse(insider_trades=trades)
+
+    @app.get("/api/market/status")
+    async def api_get_market_status() -> dict[str, Any]:
+        """Return current market status using the existing market service logic."""
+        service = MarketService(tickers=[])
+        return service.get_market_status()
+
+    @app.get("/api/market-cap")
+    async def api_get_market_cap(
+        ticker: str = Query(..., min_length=1),
+        end_date: str = Query(...),
+    ) -> dict[str, Any]:
+        """Return market cap for one ticker/date."""
+        market_cap = get_market_cap(ticker=ticker, end_date=end_date)
+        return {
+            "ticker": ticker,
+            "end_date": end_date,
+            "market_cap": market_cap,
+        }
+
+    @app.get("/api/line-items", response_model=LineItemResponse)
+    async def api_get_line_items(
+        ticker: str = Query(..., min_length=1),
+        line_items: list[str] = Query(...),
+        end_date: str = Query(...),
+        period: str = Query("ttm"),
+        limit: int = Query(10, ge=1, le=100),
+    ) -> LineItemResponse:
+        items = search_line_items(
+            ticker=ticker,
+            line_items=line_items,
+            end_date=end_date,
+            period=period,
+            limit=limit,
+        )
+        return LineItemResponse(search_results=items)
+
+    return app
+
+
+app = create_app()
+
+
+if __name__ == "__main__":
+    import uvicorn
+
+    uvicorn.run(app, host="0.0.0.0", port=8001)
--- a/backend/cli.py
+++ b/backend/cli.py
@@ -8,27 +8,62 @@ and frontend development server.
 """
 # flake8: noqa: E501
 # pylint: disable=R0912, R0915
+import logging
 import os
 import shutil
 import subprocess
 import sys
-from datetime import datetime
+from datetime import datetime, timedelta
 from pathlib import Path
 from typing import Optional
 from zoneinfo import ZoneInfo

+logger = logging.getLogger(__name__)
+
 import typer
+import yaml
 from rich.console import Console
 from rich.panel import Panel
 from rich.prompt import Confirm
+from rich.table import Table
+from dotenv import load_dotenv
+
+from backend.agents.agent_workspace import load_agent_workspace_config
+from backend.agents.prompt_loader import get_prompt_loader
+from backend.agents.skills_manager import SkillsManager
+from backend.agents.team_pipeline_config import (
+    ensure_team_pipeline_config,
+    load_team_pipeline_config,
+)
+from backend.agents.workspace_manager import WorkspaceManager
+from backend.config.constants import ANALYST_TYPES
+from backend.data.market_ingest import ingest_symbols
+from backend.data.market_store import MarketStore
+from backend.enrich.llm_enricher import get_explain_model_info, llm_enrichment_enabled
+from backend.enrich.news_enricher import enrich_symbols

 app = typer.Typer(
    name="evotraders",
    help="EvoTraders: A self-evolving multi-agent trading system",
    add_completion=False,
 )
+ingest_app = typer.Typer(help="Ingest Polygon market data into the research warehouse.")
+app.add_typer(ingest_app, name="ingest")
+skills_app = typer.Typer(help="Inspect and manage per-agent skills.")
+app.add_typer(skills_app, name="skills")
+team_app = typer.Typer(help="Inspect and manage run-scoped team pipeline config.")
+app.add_typer(team_app, name="team")

 console = Console()
+_prompt_loader = get_prompt_loader()
+load_dotenv()
+
+
+def _normalize_typer_value(value, default):
+    """Allow CLI command functions to be called directly in tests/internal code."""
+    if hasattr(value, "default"):
+        return value.default
+    return default if value is None else value


 def get_project_root() -> Path:
@@ -45,9 +80,8 @@ def handle_history_cleanup(config_name: str, auto_clean: bool = False) -> None:
        config_name: Configuration name for the run
        auto_clean: If True, skip confirmation and clean automatically
    """
-    # logs_dir = get_project_root() / "logs"
-    logs_dir = get_project_root()
-    base_data_dir = logs_dir / config_name
+    workspace_manager = WorkspaceManager(project_root=get_project_root())
+    base_data_dir = workspace_manager.get_run_dir(config_name)

    # Check if historical data exists
    if not base_data_dir.exists() or not any(base_data_dir.iterdir()):
@@ -72,8 +106,8 @@ def handle_history_cleanup(config_name: str, auto_clean: bool = False) -> None:
            )
        else:
            console.print(f"   Directory size: [cyan]{size_mb:.1f} MB[/cyan]")
-    except Exception:
-        pass
+    except Exception as e:
+        logger.debug(f"Could not calculate directory size: {e}")

    # Show last modified time
    state_dir = base_data_dir / "state"
@@ -174,12 +208,659 @@ def run_data_updater(project_root: Path) -> None:
            console.print(
                "[yellow]  Data updater module not available, skipping update[/yellow]\n",
            )
-    except Exception:
+    except Exception as e:
+        logger.debug(f"Data updater check failed: {e}")
        console.print(
            "[yellow]  Data updater check failed, skipping update[/yellow]\n",
        )


+def initialize_workspace(config_name: str) -> Path:
+    """Create run-scoped workspace files for a config."""
+    workspace_manager = WorkspaceManager(project_root=get_project_root())
+    workspace_manager.initialize_default_assets(
+        config_name=config_name,
+        agent_ids=[
+            "fundamentals_analyst",
+            "technical_analyst",
+            "sentiment_analyst",
+            "valuation_analyst",
+            "risk_manager",
+            "portfolio_manager",
+        ],
+        analyst_personas=_prompt_loader.load_yaml_config(
+            "analyst",
+            "personas",
+        ),
+    )
+    return workspace_manager.get_run_dir(config_name)
+
+
+def _require_agent_asset_dir(config_name: str, agent_id: str) -> Path:
+    manager = WorkspaceManager(project_root=get_project_root())
+    manager.initialize_default_assets(
+        config_name=config_name,
+        agent_ids=[agent_id],
+        analyst_personas=_prompt_loader.load_yaml_config(
+            "analyst",
+            "personas",
+        ),
+    )
+    return manager.skills_manager.get_agent_asset_dir(config_name, agent_id)
+
+
+def _resolve_symbols(raw_tickers: Optional[str], config_name: Optional[str] = None) -> list[str]:
+    """Resolve symbols from explicit input or runtime bootstrap config."""
+    if raw_tickers and raw_tickers.strip():
+        return [
+            item.strip().upper()
+            for item in raw_tickers.split(",")
+            if item.strip()
+        ]
+
+    workspace_manager = WorkspaceManager(project_root=get_project_root())
+    bootstrap_path = workspace_manager.get_run_dir(config_name or "default") / "BOOTSTRAP.md"
+    if bootstrap_path.exists():
+        content = bootstrap_path.read_text(encoding="utf-8")
+        for line in content.splitlines():
+            if line.strip().startswith("tickers:"):
+                raw = line.split(":", 1)[1]
+                return [
+                    item.strip().upper()
+                    for item in raw.split(",")
+                    if item.strip()
+                ]
+    return []
+
+
+def _filter_problematic_report_rows(rows: list[dict]) -> list[dict]:
+    """Keep tickers with incomplete coverage or without any LLM-enriched rows."""
+    return [
+        row
+        for row in rows
+        if float(row.get("coverage_pct") or 0.0) < 100.0
+        or int(row.get("llm_count") or 0) == 0
+    ]
+
+
+def auto_update_market_store(
+    config_name: str,
+    *,
+    end_date: Optional[str] = None,
+) -> None:
+    """Refresh the long-lived Polygon market store for the active watchlist."""
+    api_key = os.getenv("POLYGON_API_KEY", "").strip()
+    if not api_key:
+        console.print(
+            "[dim]Skipping Polygon market store update: POLYGON_API_KEY not set[/dim]",
+        )
+        return
+
+    symbols = _resolve_symbols(None, config_name)
+    if not symbols:
+        console.print(
+            f"[dim]Skipping Polygon market store update: no tickers found for config '{config_name}'[/dim]",
+        )
+        return
+
+    target_end = end_date or datetime.now().date().isoformat()
+    console.print(
+        f"[cyan]Updating Polygon market store for {', '.join(symbols)} -> {target_end}[/cyan]",
+    )
+
+    try:
+        results = ingest_symbols(
+            symbols,
+            mode="incremental",
+            end_date=target_end,
+        )
+    except Exception as exc:
+        console.print(
+            f"[yellow]Polygon market store update failed, continuing startup: {exc}[/yellow]",
+        )
+        return
+
+    for result in results:
+        console.print(
+            "[green]"
+            f"{result['symbol']}"
+            "[/green] "
+            f"prices={result['prices']} news={result['news']} aligned={result['aligned']}"
+        )
+
+
+def auto_prepare_backtest_market_store(
+    config_name: str,
+    *,
+    start_date: str,
+    end_date: str,
+) -> None:
+    """Ensure the market store has the requested backtest window for the active watchlist."""
+    api_key = os.getenv("POLYGON_API_KEY", "").strip()
+    if not api_key:
+        console.print(
+            "[dim]Skipping Polygon backtest preload: POLYGON_API_KEY not set[/dim]",
+        )
+        return
+
+    symbols = _resolve_symbols(None, config_name)
+    if not symbols:
+        console.print(
+            f"[dim]Skipping Polygon backtest preload: no tickers found for config '{config_name}'[/dim]",
+        )
+        return
+
+    console.print(
+        f"[cyan]Preparing Polygon market store for backtest {start_date} -> {end_date} "
+        f"({', '.join(symbols)})[/cyan]",
+    )
+
+    try:
+        results = ingest_symbols(
+            symbols,
+            mode="full",
+            start_date=start_date,
+            end_date=end_date,
+        )
+    except Exception as exc:
+        console.print(
+            f"[yellow]Polygon backtest preload failed, continuing startup: {exc}[/yellow]",
+        )
+        return
+
+    for result in results:
+        console.print(
+            "[green]"
+            f"{result['symbol']}"
+            "[/green] "
+            f"prices={result['prices']} news={result['news']} aligned={result['aligned']}"
+        )
+
+
+def auto_enrich_market_store(
+    config_name: str,
+    *,
+    end_date: Optional[str] = None,
+    lookback_days: int = 120,
+    force: bool = False,
+) -> None:
+    """Refresh explain-oriented enriched news for the active watchlist."""
+    symbols = _resolve_symbols(None, config_name)
+    if not symbols:
+        console.print(
+            f"[dim]Skipping explain enrich: no tickers found for config '{config_name}'[/dim]",
+        )
+        return
+
+    target_end = end_date or datetime.now().date().isoformat()
+    try:
+        end_dt = datetime.strptime(target_end, "%Y-%m-%d")
+    except ValueError:
+        console.print(
+            f"[yellow]Skipping explain enrich: invalid end date {target_end}[/yellow]",
+        )
+        return
+
+    start_date = (end_dt - timedelta(days=max(1, lookback_days))).date().isoformat()
+    console.print(
+        f"[cyan]Refreshing explain enrich for {', '.join(symbols)} -> {target_end}[/cyan]",
+    )
+    store = MarketStore()
+    try:
+        results = enrich_symbols(
+            store,
+            symbols,
+            start_date=start_date,
+            end_date=target_end,
+            limit=300,
+            skip_existing=not force,
+        )
+    except Exception as exc:
+        console.print(
+            f"[yellow]Explain enrich failed, continuing startup: {exc}[/yellow]",
+        )
+        return
+
+    for result in results:
+        console.print(
+            "[green]"
+            f"{result['symbol']}"
+            "[/green] "
+            f"news={result['news_count']} queued={result['queued_count']} analyzed={result['analyzed']} "
+            f"skipped={result['skipped_existing_count']} deduped={result['deduped_count']} "
+            f"llm={result['llm_count']} local={result['local_count']}"
+        )
+
+
+@app.command("init-workspace")
+def init_workspace(
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Configuration name for the workspace",
+    ),
+):
+    """Initialize run-scoped BOOTSTRAP and agent prompt asset files."""
+    run_dir = initialize_workspace(config_name)
+    console.print(
+        Panel.fit(
+            f"[bold green]Workspace initialized[/bold green]\n[cyan]{run_dir}[/cyan]",
+            border_style="green",
+        ),
+    )
+
+
+@ingest_app.command("full")
+def ingest_full(
+    tickers: Optional[str] = typer.Option(
+        None,
+        "--tickers",
+        "-t",
+        help="Comma-separated tickers to ingest",
+    ),
+    start: Optional[str] = typer.Option(
+        None,
+        "--start",
+        help="Start date for full ingestion (YYYY-MM-DD)",
+    ),
+    end: Optional[str] = typer.Option(
+        None,
+        "--end",
+        help="End date for ingestion (YYYY-MM-DD)",
+    ),
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Fallback config to read tickers from BOOTSTRAP.md",
+    ),
+):
+    """Run full Polygon ingestion for the specified symbols."""
+    symbols = _resolve_symbols(tickers, config_name)
+    if not symbols:
+        console.print("[red]No tickers provided and none found in BOOTSTRAP.md[/red]")
+        raise typer.Exit(1)
+
+    console.print(f"[cyan]Starting full Polygon ingest for {', '.join(symbols)}[/cyan]")
+    results = ingest_symbols(symbols, mode="full", start_date=start, end_date=end)
+    for result in results:
+        console.print(
+            f"[green]{result['symbol']}[/green] prices={result['prices']} news={result['news']} aligned={result['aligned']}"
+        )
+
+
+@ingest_app.command("update")
+def ingest_update(
+    tickers: Optional[str] = typer.Option(
+        None,
+        "--tickers",
+        "-t",
+        help="Comma-separated tickers to update",
+    ),
+    end: Optional[str] = typer.Option(
+        None,
+        "--end",
+        help="Optional end date override (YYYY-MM-DD)",
+    ),
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Fallback config to read tickers from BOOTSTRAP.md",
+    ),
+):
+    """Run incremental Polygon ingestion using stored watermarks."""
+    symbols = _resolve_symbols(tickers, config_name)
+    if not symbols:
+        console.print("[red]No tickers provided and none found in BOOTSTRAP.md[/red]")
+        raise typer.Exit(1)
+
+    console.print(f"[cyan]Starting incremental Polygon ingest for {', '.join(symbols)}[/cyan]")
+    results = ingest_symbols(symbols, mode="incremental", end_date=end)
+    for result in results:
+        console.print(
+            f"[green]{result['symbol']}[/green] prices={result['prices']} news={result['news']} aligned={result['aligned']}"
+        )
+
+
+@ingest_app.command("enrich")
+def ingest_enrich(
+    tickers: Optional[str] = typer.Option(
+        None,
+        "--tickers",
+        "-t",
+        help="Comma-separated tickers to enrich",
+    ),
+    start: Optional[str] = typer.Option(
+        None,
+        "--start",
+        help="Optional start date for enrichment window (YYYY-MM-DD)",
+    ),
+    end: Optional[str] = typer.Option(
+        None,
+        "--end",
+        help="Optional end date for enrichment window (YYYY-MM-DD)",
+    ),
+    limit: int = typer.Option(
+        300,
+        "--limit",
+        help="Maximum raw news rows per ticker to analyze",
+    ),
+    force: bool = typer.Option(
+        False,
+        "--force",
+        help="Re-analyze already enriched news instead of only missing rows",
+    ),
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Fallback config to read tickers from BOOTSTRAP.md",
+    ),
+):
+    """Run explain-oriented news enrichment for symbols already in the market store."""
+    symbols = _resolve_symbols(tickers, config_name)
+    if not symbols:
+        console.print("[red]No tickers provided and none found in BOOTSTRAP.md[/red]")
+        raise typer.Exit(1)
+
+    console.print(f"[cyan]Starting explain enrich for {', '.join(symbols)}[/cyan]")
+    store = MarketStore()
+    results = enrich_symbols(
+        store,
+        symbols,
+        start_date=start,
+        end_date=end,
+        limit=max(10, limit),
+        skip_existing=not force,
+    )
+    for result in results:
+        console.print(
+            f"[green]{result['symbol']}[/green] "
+            f"news={result['news_count']} queued={result['queued_count']} analyzed={result['analyzed']} "
+            f"skipped={result['skipped_existing_count']} deduped={result['deduped_count']} "
+            f"llm={result['llm_count']} local={result['local_count']}"
+        )
+
+
+@ingest_app.command("report")
+def ingest_report(
+    tickers: Optional[str] = typer.Option(
+        None,
+        "--tickers",
+        "-t",
+        help="Optional comma-separated tickers to report",
+    ),
+    start: Optional[str] = typer.Option(
+        None,
+        "--start",
+        help="Optional start date for report window (YYYY-MM-DD)",
+    ),
+    end: Optional[str] = typer.Option(
+        None,
+        "--end",
+        help="Optional end date for report window (YYYY-MM-DD)",
+    ),
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Fallback config to read tickers from BOOTSTRAP.md",
+    ),
+    only_problematic: bool = typer.Option(
+        False,
+        "--only-problematic",
+        help="Only show tickers with incomplete coverage or no LLM-enriched news",
+    ),
+):
+    """Show explain enrichment coverage and freshness per ticker."""
+    symbols = _resolve_symbols(tickers, config_name)
+    store = MarketStore()
+    report_rows = store.get_enrich_report(
+        symbols=symbols or None,
+        start_date=start,
+        end_date=end,
+    )
+    if only_problematic:
+        report_rows = _filter_problematic_report_rows(report_rows)
+    if not report_rows:
+        if only_problematic:
+            console.print("[green]No problematic enrich report rows found for the requested scope[/green]")
+        else:
+            console.print("[yellow]No enrich report rows found for the requested scope[/yellow]")
+        raise typer.Exit(0)
+
+    model_info = get_explain_model_info()
+    model_label = model_info["label"] if llm_enrichment_enabled() else "disabled"
+    table = Table(title="Explain Enrichment Report")
+    table.add_column("Ticker", style="cyan")
+    table.add_column("Raw News", justify="right")
+    table.add_column("Analyzed", justify="right")
+    table.add_column("Coverage", justify="right")
+    table.add_column("LLM", justify="right")
+    table.add_column("Local", justify="right")
+    table.add_column("Latest Trade Date")
+    table.add_column("Latest Analysis")
+    table.caption = f"Explain LLM: {model_label}"
+
+    for row in report_rows:
+        table.add_row(
+            row["symbol"],
+            str(row["raw_news_count"]),
+            str(row["analyzed_news_count"]),
+            f'{row["coverage_pct"]:.1f}%',
+            str(row["llm_count"]),
+            str(row["local_count"]),
+            str(row["latest_trade_date"] or "-"),
+            str(row["latest_analysis_at"] or "-"),
+        )
+    console.print(table)
+
+
+@skills_app.command("list")
+def skills_list(
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Run config name.",
+    ),
+    agent_id: Optional[str] = typer.Option(
+        None,
+        "--agent-id",
+        "-a",
+        help="Optional agent id to show resolved status for.",
+    ),
+):
+    """List available skills and optional agent-level enablement state."""
+    project_root = get_project_root()
+    skills_manager = SkillsManager(project_root=project_root)
+    catalog = (
+        skills_manager.list_agent_skill_catalog(config_name, agent_id)
+        if agent_id
+        else skills_manager.list_skill_catalog()
+    )
+    if not catalog:
+        console.print("[yellow]No skills found[/yellow]")
+        raise typer.Exit(0)
+
+    agent_config = None
+    resolved_skills = set()
+    if agent_id:
+        asset_dir = _require_agent_asset_dir(config_name, agent_id)
+        agent_config = load_agent_workspace_config(asset_dir / "agent.yaml")
+        resolved_skills = set(
+            skills_manager.resolve_agent_skill_names(
+                config_name=config_name,
+                agent_id=agent_id,
+                default_skills=[],
+            ),
+        )
+
+    table = Table(title="Skill Catalog")
+    table.add_column("Skill", style="cyan")
+    table.add_column("Source")
+    table.add_column("Description")
+    if agent_id:
+        table.add_column("Status")
+
+    enabled = set(agent_config.enabled_skills) if agent_config else set()
+    disabled = set(agent_config.disabled_skills) if agent_config else set()
+    for skill in catalog:
+        row = [
+            skill.skill_name,
+            skill.source,
+            skill.description or "-",
+        ]
+        if agent_id:
+            if skill.skill_name in disabled:
+                status = "disabled"
+            elif skill.skill_name in enabled:
+                status = "enabled"
+            elif skill.skill_name in resolved_skills:
+                status = "active"
+            else:
+                status = "-"
+            row.append(status)
+        table.add_row(*row)
+    console.print(table)
+
+
+@skills_app.command("enable")
+def skills_enable(
+    agent_id: str = typer.Option(..., "--agent-id", "-a", help="Agent id."),
+    skill: str = typer.Option(..., "--skill", "-s", help="Skill name."),
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Run config name.",
+    ),
+):
+    """Enable a skill for one agent in agent.yaml."""
+    asset_dir = _require_agent_asset_dir(config_name, agent_id)
+    skills_manager = SkillsManager(project_root=get_project_root())
+    catalog = {
+        item.skill_name
+        for item in skills_manager.list_agent_skill_catalog(config_name, agent_id)
+    }
+    if skill not in catalog:
+        console.print(f"[red]Unknown skill: {skill}[/red]")
+        raise typer.Exit(1)
+
+    result = skills_manager.update_agent_skill_overrides(
+        config_name=config_name,
+        agent_id=agent_id,
+        enable=[skill],
+    )
+    console.print(
+        f"[green]Enabled[/green] `{skill}` for `{agent_id}` "
+        f"([{asset_dir / 'agent.yaml'}])",
+    )
+    console.print(f"Enabled skills: {', '.join(result['enabled_skills']) or '-'}")
+    console.print(f"Disabled skills: {', '.join(result['disabled_skills']) or '-'}")
+
+
+@skills_app.command("disable")
+def skills_disable(
+    agent_id: str = typer.Option(..., "--agent-id", "-a", help="Agent id."),
+    skill: str = typer.Option(..., "--skill", "-s", help="Skill name."),
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Run config name.",
+    ),
+):
+    """Disable a skill for one agent in agent.yaml."""
+    asset_dir = _require_agent_asset_dir(config_name, agent_id)
+    skills_manager = SkillsManager(project_root=get_project_root())
+    result = skills_manager.update_agent_skill_overrides(
+        config_name=config_name,
+        agent_id=agent_id,
+        disable=[skill],
+    )
+    console.print(
+        f"[yellow]Disabled[/yellow] `{skill}` for `{agent_id}` "
+        f"([{asset_dir / 'agent.yaml'}])",
+    )
+    console.print(f"Enabled skills: {', '.join(result['enabled_skills']) or '-'}")
+    console.print(f"Disabled skills: {', '.join(result['disabled_skills']) or '-'}")
+
+
+@skills_app.command("install")
+def skills_install(
+    agent_id: str = typer.Option(..., "--agent-id", "-a", help="Target agent id."),
+    source: str = typer.Option(
+        ...,
+        "--source",
+        "-s",
+        help="External skill source: directory path, zip path, or http(s) zip URL.",
+    ),
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Run config name.",
+    ),
+    name: Optional[str] = typer.Option(
+        None,
+        "--name",
+        help="Optional override skill name.",
+    ),
+    activate: bool = typer.Option(
+        True,
+        "--activate/--no-activate",
+        help="Enable the skill for this agent immediately.",
+    ),
+):
+    """Install an external skill into one agent's local skill directory."""
+    _require_agent_asset_dir(config_name, agent_id)
+    skills_manager = SkillsManager(project_root=get_project_root())
+    result = skills_manager.install_external_skill_for_agent(
+        config_name=config_name,
+        agent_id=agent_id,
+        source=source,
+        skill_name=name,
+        activate=activate,
+    )
+    console.print(
+        f"[green]Installed[/green] `{result['skill_name']}` to `{agent_id}`",
+    )
+    console.print(f"Path: {result['target_dir']}")
+    console.print(f"Activated: {result['activated']}")
+    warnings = result.get("warnings") or []
+    if warnings:
+        console.print(f"Warnings: {'; '.join(warnings)}")
+
+
+@team_app.command("show")
+def team_show(
+    config_name: str = typer.Option(
+        "default",
+        "--config-name",
+        "-c",
+        help="Run config name.",
+    ),
+):
+    """Show TEAM_PIPELINE.yaml for one run."""
+    project_root = get_project_root()
+    ensure_team_pipeline_config(
+        project_root=project_root,
+        config_name=config_name,
+        default_analysts=list(ANALYST_TYPES.keys()),
+    )
+    config = load_team_pipeline_config(project_root, config_name)
+    console.print(
+        Panel.fit(
+            yaml.safe_dump(config, allow_unicode=True, sort_keys=False),
+            title=f"TEAM_PIPELINE ({config_name})",
+            border_style="cyan",
+        ),
+    )
+
+
@app.command()
 def backtest(
    start: Optional[str] = typer.Option(
@@ -242,6 +923,7 @@ def backtest(
            border_style="cyan",
        ),
    )
+    poll_interval = int(_normalize_typer_value(poll_interval, 10))

    # Validate dates - required for backtest
    if not start or not end:
@@ -288,6 +970,16 @@ def backtest(

    # Run data updater
    run_data_updater(project_root)
+    auto_prepare_backtest_market_store(
+        config_name,
+        start_date=start,
+        end_date=end,
+    )
+    auto_enrich_market_store(
+        config_name,
+        end_date=end,
+        force=False,
+    )

    # Build command using backend.main
    cmd = [
@@ -349,12 +1041,22 @@ def live(
        "-p",
        help="WebSocket server port",
    ),
+    schedule_mode: str = typer.Option(
+        "daily",
+        "--schedule-mode",
+        help="Scheduler mode: 'daily' or 'intraday'",
+    ),
    trigger_time: str = typer.Option(
        "now",
        "--trigger-time",
        "-t",
        help="Trigger time in LOCAL timezone (HH:MM), or 'now' to run immediately",
    ),
+    interval_minutes: int = typer.Option(
+        60,
+        "--interval-minutes",
+        help="When schedule-mode=intraday, run every N minutes",
+    ),
    poll_interval: int = typer.Option(
        10,
        "--poll-interval",
@@ -378,9 +1080,12 @@ def live(
        evotraders live                    # Run immediately (default)
        evotraders live --mock             # Mock mode
        evotraders live -t 22:30           # Run at 22:30 local time daily
+        evotraders live --schedule-mode intraday --interval-minutes 60
        evotraders live --trigger-time now # Run immediately
        evotraders live --clean            # Clear historical data before starting
    """
+    schedule_mode = str(_normalize_typer_value(schedule_mode, "daily"))
+    interval_minutes = int(_normalize_typer_value(interval_minutes, 60))
    mode_name = "MOCK" if mock else "LIVE"
    console.print(
        Panel.fit(
@@ -412,6 +1117,16 @@ def live(
    # Handle historical data cleanup
    handle_history_cleanup(config_name, auto_clean=clean)

+    if schedule_mode not in {"daily", "intraday"}:
+        console.print(
+            f"[red]Error: unsupported schedule mode '{schedule_mode}'[/red]",
+        )
+        raise typer.Exit(1)
+
+    if interval_minutes <= 0:
+        console.print("[red]Error: --interval-minutes must be > 0[/red]")
+        raise typer.Exit(1)
+
    # Convert local time to NYSE time
    nyse_tz = ZoneInfo("America/New_York")
    local_tz = datetime.now().astimezone().tzinfo
@@ -419,7 +1134,9 @@ def live(
    nyse_now = datetime.now(nyse_tz)

    # Convert trigger time from local to NYSE
-    if trigger_time.lower() == "now":
+    if schedule_mode == "intraday":
+        nyse_trigger_time = "now"
+    elif trigger_time.lower() == "now":
        nyse_trigger_time = "now"
    else:
        local_trigger = datetime.strptime(trigger_time, "%H:%M")
@@ -439,7 +1156,10 @@ def live(
    console.print(
        f"   NYSE Time:  {nyse_now.strftime('%Y-%m-%d %H:%M:%S %Z')}",
    )
-    if nyse_trigger_time == "now":
+    console.print(f"   Schedule:   {schedule_mode}")
+    if schedule_mode == "intraday":
+        console.print(f"   Interval:   every {interval_minutes} minute(s)")
+    elif nyse_trigger_time == "now":
        console.print("   Trigger:    [green]NOW (immediate)[/green]")
    else:
        console.print(
@@ -471,6 +1191,15 @@ def live(
    # Data update (if not mock mode)
    if not mock:
        run_data_updater(project_root)
+        auto_update_market_store(
+            config_name,
+            end_date=nyse_now.date().isoformat(),
+        )
+        auto_enrich_market_store(
+            config_name,
+            end_date=nyse_now.date().isoformat(),
+            force=False,
+        )
    else:
        console.print(
            "\n[dim]Mock mode enabled - skipping data update[/dim]\n",
@@ -490,10 +1219,14 @@ def live(
        host,
        "--port",
        str(port),
+        "--schedule-mode",
+        schedule_mode,
        "--poll-interval",
        str(poll_interval),
        "--trigger-time",
        nyse_trigger_time,
+        "--interval-minutes",
+        str(interval_minutes),
    ]

    if mock:
--- a/backend/config/agent_profiles.yaml
+++ b/backend/config/agent_profiles.yaml
@@ -0,0 +1,37 @@
+fundamentals_analyst:
+  skills:
+    - fundamental_review
+  active_tool_groups:
+    - fundamentals
+    - valuation
+
+technical_analyst:
+  skills:
+    - technical_review
+  active_tool_groups:
+    - technical
+
+sentiment_analyst:
+  skills:
+    - sentiment_review
+  active_tool_groups:
+    - sentiment
+
+valuation_analyst:
+  skills:
+    - valuation_review
+  active_tool_groups:
+    - valuation
+    - fundamentals
+
+portfolio_manager:
+  skills:
+    - portfolio_decisioning
+  active_tool_groups:
+    - portfolio_ops
+
+risk_manager:
+  skills:
+    - risk_review
+  active_tool_groups:
+    - risk_ops
--- a/backend/config/bootstrap_config.py
+++ b/backend/config/bootstrap_config.py
@@ -0,0 +1,163 @@
+# -*- coding: utf-8 -*-
+"""Parse run-scoped BOOTSTRAP.md into structured and runtime config."""
+
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Dict
+import re
+
+import yaml
+
+from backend.config.env_config import get_env_float, get_env_int, get_env_list
+
+
+BOOTSTRAP_FRONT_MATTER_RE = re.compile(
+    r"^---\s*\n(.*?)\n---\s*\n?(.*)$",
+    re.DOTALL,
+)
+
+
+@dataclass(frozen=True)
+class BootstrapConfig:
+    """Structured configuration extracted from BOOTSTRAP.md."""
+
+    values: Dict[str, Any] = field(default_factory=dict)
+    prompt_body: str = ""
+
+    def get(self, key: str, default: Any = None) -> Any:
+        return self.values.get(key, default)
+
+    def agent_override(self, agent_id: str) -> Dict[str, Any]:
+        overrides = self.values.get("agent_overrides", {})
+        if not isinstance(overrides, dict):
+            return {}
+        override = overrides.get(agent_id, {})
+        return override if isinstance(override, dict) else {}
+
+
+def load_bootstrap_config(bootstrap_path: Path) -> BootstrapConfig:
+    """Load structured bootstrap config and free-form prompt body."""
+    if not bootstrap_path.exists():
+        return BootstrapConfig()
+
+    raw = bootstrap_path.read_text(encoding="utf-8").strip()
+    if not raw:
+        return BootstrapConfig()
+
+    match = BOOTSTRAP_FRONT_MATTER_RE.match(raw)
+    if not match:
+        return BootstrapConfig(prompt_body=raw)
+
+    front_matter = match.group(1).strip()
+    body = match.group(2).strip()
+    parsed = yaml.safe_load(front_matter) or {}
+    if not isinstance(parsed, dict):
+        parsed = {}
+
+    return BootstrapConfig(values=parsed, prompt_body=body)
+
+
+def get_bootstrap_config_for_run(
+    project_root: Path,
+    config_name: str,
+) -> BootstrapConfig:
+    """Load BOOTSTRAP.md from the run workspace."""
+    return load_bootstrap_config(
+        project_root / "runs" / config_name / "BOOTSTRAP.md",
+    )
+
+
+def save_bootstrap_config(bootstrap_path: Path, config: BootstrapConfig) -> None:
+    """Persist structured bootstrap config back to BOOTSTRAP.md."""
+    bootstrap_path.parent.mkdir(parents=True, exist_ok=True)
+    values = config.values if isinstance(config.values, dict) else {}
+    front_matter = yaml.safe_dump(
+        values,
+        allow_unicode=True,
+        sort_keys=False,
+    ).strip()
+    body = (config.prompt_body or "").strip()
+
+    content = f"---\n{front_matter}\n---"
+    if body:
+        content += f"\n\n{body}\n"
+    else:
+        content += "\n"
+
+    bootstrap_path.write_text(content, encoding="utf-8")
+
+
+def update_bootstrap_values_for_run(
+    project_root: Path,
+    config_name: str,
+    updates: Dict[str, Any],
+) -> BootstrapConfig:
+    """Patch selected front matter keys for a run and persist them."""
+    bootstrap_path = project_root / "runs" / config_name / "BOOTSTRAP.md"
+    existing = load_bootstrap_config(bootstrap_path)
+    values = dict(existing.values)
+    values.update(updates)
+    updated = BootstrapConfig(values=values, prompt_body=existing.prompt_body)
+    save_bootstrap_config(bootstrap_path, updated)
+    return updated
+
+
+def _coerce_bool(value: Any) -> bool:
+    """Parse booleans from bootstrap-friendly string values."""
+    if isinstance(value, bool):
+        return value
+    if isinstance(value, str):
+        normalized = value.strip().lower()
+        if normalized in {"1", "true", "yes", "on"}:
+            return True
+        if normalized in {"0", "false", "no", "off"}:
+            return False
+    return bool(value)
+
+
+def resolve_runtime_config(
+    project_root: Path,
+    config_name: str,
+    enable_memory: bool = False,
+    schedule_mode: str = "daily",
+    interval_minutes: int = 60,
+    trigger_time: str = "09:30",
+) -> Dict[str, Any]:
+    """Merge env defaults with run-scoped bootstrap front matter."""
+    bootstrap = get_bootstrap_config_for_run(project_root, config_name)
+    return {
+        "tickers": bootstrap.get("tickers")
+        or get_env_list("TICKERS", ["AAPL", "MSFT"]),
+        "initial_cash": float(
+            bootstrap.get(
+                "initial_cash",
+                get_env_float("INITIAL_CASH", 100000.0),
+            ),
+        ),
+        "margin_requirement": float(
+            bootstrap.get(
+                "margin_requirement",
+                get_env_float("MARGIN_REQUIREMENT", 0.0),
+            ),
+        ),
+        "max_comm_cycles": int(
+            bootstrap.get(
+                "max_comm_cycles",
+                get_env_int("MAX_COMM_CYCLES", 2),
+            ),
+        ),
+        "schedule_mode": str(
+            bootstrap.get("schedule_mode", schedule_mode),
+        ).strip().lower() or schedule_mode,
+        "interval_minutes": int(
+            bootstrap.get(
+                "interval_minutes",
+                interval_minutes or get_env_int("INTERVAL_MINUTES", 60),
+            ),
+        ),
+        "trigger_time": str(
+            bootstrap.get("trigger_time", trigger_time),
+        ).strip() or trigger_time,
+        "enable_memory": bool(enable_memory)
+        or _coerce_bool(bootstrap.get("enable_memory", False)),
+    }
--- a/backend/config/env_config.py
+++ b/backend/config/env_config.py
@@ -3,6 +3,7 @@
 """Environment config helpers with light validation and normalization."""

 import os
+import warnings
 from dataclasses import dataclass
 from typing import Optional

@@ -16,6 +17,36 @@ PROVIDER_ALIASES = {
    "vertexai": "GEMINI",
 }

+# Default dev CORS origins (localhost variants used by common dev servers)
+_LOCALHOST_ORIGINS = [
+    "http://localhost:5173",
+    "http://localhost:3000",
+    "http://localhost:8000",
+    "http://127.0.0.1:5173",
+    "http://127.0.0.1:3000",
+    "http://127.0.0.1:8000",
+]
+
+
+def get_cors_origins() -> list[str]:
+    """Get CORS allowed origins from environment.
+
+    Reads CORS_ALLOWED_ORIGINS env var (comma-separated).
+    Falls back to localhost dev origins if not set.
+    Warns if "*" is configured (only acceptable for local dev).
+    """
+    origins = get_env_list("CORS_ALLOWED_ORIGINS", default=[])
+    if origins:
+        if "*" in origins:
+            warnings.warn(
+                "CORS_ALLOWED_ORIGINS contains '*' — this allows any origin. "
+                "Only use in local development, never in production.",
+                UserWarning,
+            )
+        return origins
+    # Fallback: local dev only
+    return _LOCALHOST_ORIGINS
+

@dataclass(frozen=True)
 class AgentModelConfig:
--- a/backend/core/init.py
+++ b/backend/core/init.py
@@ -1,7 +1,35 @@
 # -*- coding: utf-8 -*-
-"""Core pipeline and orchestration logic"""
+"""Core pipeline and orchestration logic.
+
+Keep ``pipeline_runner`` behind lazy wrappers so importing ``backend.core`` does
+not immediately pull in the gateway runtime graph.
+"""

 from .pipeline import TradingPipeline
 from .state_sync import StateSync

-__all__ = ["TradingPipeline", "StateSync"]
+
+def create_agents(*args, **kwargs):
+    from .pipeline_runner import create_agents as _create_agents
+
+    return _create_agents(*args, **kwargs)
+
+
+def create_long_term_memory(*args, **kwargs):
+    from .pipeline_runner import create_long_term_memory as _create_long_term_memory
+
+    return _create_long_term_memory(*args, **kwargs)
+
+
+def stop_gateway(*args, **kwargs):
+    from .pipeline_runner import stop_gateway as _stop_gateway
+
+    return _stop_gateway(*args, **kwargs)
+
+__all__ = [
+    "TradingPipeline",
+    "StateSync",
+    "create_agents",
+    "create_long_term_memory",
+    "stop_gateway",
+]
--- a/backend/core/pipeline.py
+++ b/backend/core/pipeline.py
@@ -10,6 +10,8 @@ import json
 import logging
 import os
 import re
+from contextlib import nullcontext
+from pathlib import Path
 from typing import Any, Awaitable, Callable, Dict, List, Optional

 from agentscope.message import Msg
@@ -19,6 +21,28 @@ from backend.utils.settlement import SettlementCoordinator
 from backend.utils.terminal_dashboard import get_dashboard
 from backend.core.state_sync import StateSync
 from backend.utils.trade_executor import PortfolioTradeExecutor
+from backend.runtime.manager import TradingRuntimeManager
+from backend.runtime.session import TradingSessionKey
+from backend.agents.team_pipeline_config import (
+    resolve_active_analysts,
+    update_active_analysts,
+)
+from backend.agents import AnalystAgent
+from backend.agents.toolkit_factory import create_agent_toolkit
+from backend.agents.workspace_manager import WorkspaceManager
+from backend.agents.prompt_loader import get_prompt_loader
+from backend.llm.models import get_agent_formatter, get_agent_model
+from backend.config.constants import ANALYST_TYPES
+
+# Team infrastructure imports (graceful import - may not exist yet)
+try:
+    from backend.agents.team.team_coordinator import TeamCoordinator
+    from backend.agents.team.msg_hub import MsgHub as TeamMsgHub
+    TEAM_COORD_AVAILABLE = True
+except ImportError:
+    TEAM_COORD_AVAILABLE = False
+    TeamCoordinator = None
+    TeamMsgHub = None


 logger = logging.getLogger(__name__)
@@ -46,6 +70,8 @@ class TradingPipeline:
    6. Reflection phase: broadcast closing P&L, agents record to long-term memory

    Real-time updates via StateSync after each agent completes.
+
+    Supports both legacy agent lists and new workspace-based agent loading.
    """

    def __init__(
@@ -56,6 +82,9 @@ class TradingPipeline:
        state_sync: Optional["StateSync"] = None,
        settlement_coordinator: Optional[SettlementCoordinator] = None,
        max_comm_cycles: Optional[int] = None,
+        workspace_id: Optional[str] = None,
+        agent_factory: Optional[Any] = None,
+        runtime_manager: Optional[TradingRuntimeManager] = None,
    ):
        self.analysts = analysts
        self.risk_manager = risk_manager
@@ -66,6 +95,17 @@ class TradingPipeline:
            os.getenv("MAX_COMM_CYCLES", "2"),
        )
        self.conference_summary = None  # Store latest conference summary
+        self.workspace_id = workspace_id
+        self.agent_factory = agent_factory
+        self.runtime_manager = runtime_manager
+        self._session_key: Optional[str] = None
+        self._dynamic_analysts: Dict[str, Any] = {}
+
+        if hasattr(self.pm, "set_team_controller"):
+            self.pm.set_team_controller(
+                create_agent_callback=self._create_runtime_analyst,
+                remove_agent_callback=self._remove_runtime_analyst,
+            )

    async def run_cycle(
        self,
@@ -80,6 +120,7 @@ class TradingPipeline:
        get_close_prices_fn: Optional[
            Callable[[], Awaitable[Dict[str, float]]]
        ] = None,
+        execute_decisions: bool = True,
    ) -> Dict[str, Any]:
        """
        Run one complete trading cycle
@@ -101,12 +142,19 @@ class TradingPipeline:
        Each agent's result is broadcast immediately via StateSync.
        """
        _log(f"Starting cycle {date} - {len(tickers)} tickers")
+        session_key = TradingSessionKey(date=date).key()
+        self._session_key = session_key
+        active_analysts = self._get_active_analysts()
+        if self.runtime_manager:
+            self.runtime_manager.set_session_key(session_key)
+            self._runtime_log_event("cycle:start", {"tickers": tickers, "date": date})
+            self._runtime_batch_status(active_analysts, "analysis_in_progress")

        # Phase 0: Clear short-term memory to avoid cross-day context pollution
        _log("Phase 0: Clearing memory")
        await self._clear_all_agent_memory()

-        participants = self.analysts + [self.risk_manager, self.pm]
+        participants = self._all_analysts() + [self.risk_manager, self.pm]

        # Single MsgHub for entire cycle - no nesting
        async with MsgHub(
@@ -117,12 +165,17 @@ class TradingPipeline:
                "system",
            ),
        ):
-            # Phase 1.1: Analysts
-            _log("Phase 1.1: Analyst analysis")
-            analyst_results = await self._run_analysts_with_sync(tickers, date)
+            # Phase 1.1: Analysts (parallel execution with TeamCoordinator)
+            _log("Phase 1.1: Analyst analysis (parallel)")
+            analyst_results = await self._run_analysts_parallel(
+                tickers,
+                date,
+                active_analysts=active_analysts,
+            )

            # Phase 1.2: Risk Manager
            _log("Phase 1.2: Risk assessment")
+            self._runtime_update_status(self.risk_manager, "risk_assessment")
            risk_assessment = await self._run_risk_manager_with_sync(
                tickers,
                date,
@@ -145,6 +198,7 @@ class TradingPipeline:
            final_predictions = await self._collect_final_predictions(
                tickers,
                date,
+                active_analysts=active_analysts,
            )

            # Record final predictions for leaderboard ranking
@@ -161,6 +215,7 @@ class TradingPipeline:

            # Phase 3: PM makes decisions
            _log("Phase 3.1: PM makes decisions")
+            self._runtime_update_status(self.pm, "decision_phase")
            pm_result = await self._run_pm_with_sync(
                tickers,
                date,
@@ -169,10 +224,17 @@ class TradingPipeline:
                risk_assessment,
            )

-        # Phase 4: Execute decisions
-        _log("Phase 4: Executing trades")
        decisions = pm_result.get("decisions", {})
-        execution_result = self._execute_decisions(decisions, prices, date)
+        execution_result = {
+            "executed_trades": [],
+            "portfolio": self.pm.get_portfolio_state(),
+        }
+        if execute_decisions:
+            _log("Phase 4: Executing trades")
+            self._runtime_update_status(self.pm, "executing")
+            execution_result = self._execute_decisions(decisions, prices, date)
+        else:
+            _log("Phase 4: Skipping trade execution")

        # Live mode: wait for market close before settlement
        if get_close_prices_fn:
@@ -184,6 +246,10 @@ class TradingPipeline:
        settlement_result = None
        if close_prices and self.settlement_coordinator:
            _log("Phase 5: Daily review and generate memories")
+            self._runtime_batch_status(
+                [self.risk_manager] + self._all_analysts() + [self.pm],
+                "settlement",
+            )

            agent_trajectories = await self._capture_agent_trajectories()

@@ -214,8 +280,17 @@ class TradingPipeline:
                settlement_result=settlement_result,
                conference_summary=self.conference_summary,
            )
+            self._runtime_batch_status(
+                [self.risk_manager] + self._all_analysts() + [self.pm],
+                "reflection",
+            )

        _log(f"Cycle complete: {date}")
+        self._runtime_batch_status(
+            self._all_analysts() + [self.risk_manager, self.pm],
+            "idle",
+        )
+        self._runtime_log_event("cycle:end", {"tickers": tickers, "date": date})

        return {
            "analyst_results": analyst_results,
@@ -226,9 +301,54 @@ class TradingPipeline:
            "settlement_result": settlement_result,
        }

+    def reload_runtime_assets(
+        self,
+        runtime_config: Optional[Dict[str, Any]] = None,
+    ) -> Dict[str, Any]:
+        """Reload prompt assets and safe in-process runtime settings."""
+        from backend.agents.skills_manager import SkillsManager
+        from backend.agents.toolkit_factory import load_agent_profiles
+
+        config_name = getattr(self.pm, "config", {}).get("config_name", "default")
+        if runtime_config and "max_comm_cycles" in runtime_config:
+            self.max_comm_cycles = int(runtime_config["max_comm_cycles"])
+
+        skills_manager = SkillsManager()
+        profiles = load_agent_profiles()
+        active_skill_map = skills_manager.prepare_active_skills(
+            config_name=config_name,
+            agent_defaults={
+                agent_id: profile.get("skills", [])
+                for agent_id, profile in profiles.items()
+            },
+        )
+
+        for analyst in self._all_analysts():
+            analyst.reload_runtime_assets(
+                active_skill_dirs=active_skill_map.get(analyst.name, []),
+            )
+
+        self.risk_manager.reload_runtime_assets(
+            active_skill_dirs=active_skill_map.get("risk_manager", []),
+        )
+        self.pm.reload_runtime_assets(
+            active_skill_dirs=active_skill_map.get("portfolio_manager", []),
+        )
+
+        return {
+            "config_name": config_name,
+            "reloaded_agents": [agent.name for agent in self._all_analysts()]
+            + ["risk_manager", "portfolio_manager"],
+            "active_skills": {
+                agent_id: [path.name for path in paths]
+                for agent_id, paths in active_skill_map.items()
+            },
+            "max_comm_cycles": self.max_comm_cycles,
+        }
+
    async def _clear_all_agent_memory(self):
        """Clear short-term memory for all agents"""
-        for analyst in self.analysts:
+        for analyst in self._all_analysts():
            await analyst.memory.clear()

        await self.risk_manager.memory.clear()
@@ -310,7 +430,7 @@ class TradingPipeline:
        trajectories = {}

        # Capture analyst trajectories
-        for analyst in self.analysts:
+        for analyst in self._all_analysts():
            try:
                msgs = await analyst.memory.get_memory()
                if msgs:
@@ -520,7 +640,7 @@ class TradingPipeline:
        )

        # Record for analysts
-        for analyst in self.analysts:
+        for analyst in self._all_analysts():
            if (
                hasattr(analyst, "long_term_memory")
                and analyst.long_term_memory is not None
@@ -639,67 +759,82 @@ class TradingPipeline:
                date=date,
            )

-        # Run discussion cycles (no new MsgHub - use parent's)
-        for cycle in range(self.max_comm_cycles):
+        # Conference participants: analysts + PM
+        conference_participants = self._get_active_analysts() + [self.pm]
+
+        # Use TeamMsgHub for conference if available
+        if TEAM_COORD_AVAILABLE and TeamMsgHub is not None:
            _log(
-                "Phase 2.1: Conference discussion - "
-                f"Conference {cycle + 1}/{self.max_comm_cycles}",
+                f"Phase 2.1: Conference using TeamMsgHub with "
+                f"{len(conference_participants)} participants"
            )
+            conference_hub = TeamMsgHub(participants=conference_participants)
+        else:
+            _log("Phase 2.1: Conference using standard MsgHub context")
+            conference_hub = None

-            if self.state_sync:
-                await self.state_sync.on_conference_cycle_start(
-                    cycle=cycle + 1,
-                    total_cycles=self.max_comm_cycles,
+        # Run discussion cycles
+        async with conference_hub if conference_hub else nullcontext(None):
+            for cycle in range(self.max_comm_cycles):
+                _log(
+                    "Phase 2.1: Conference discussion - "
+                    f"Conference {cycle + 1}/{self.max_comm_cycles}",
                )

-            # PM sets agenda or asks questions
-            pm_prompt = self._build_pm_discussion_prompt(
-                cycle=cycle,
-                tickers=tickers,
-                date=date,
-                prices=prices,
-                analyst_results=analyst_results,
-                risk_assessment=risk_assessment,
-            )
+                if self.state_sync:
+                    await self.state_sync.on_conference_cycle_start(
+                        cycle=cycle + 1,
+                        total_cycles=self.max_comm_cycles,
+                    )

-            pm_msg = Msg(name="system", content=pm_prompt, role="user")
-            pm_response = await self.pm.reply(pm_msg)
-
-            if self.state_sync:
-                pm_content = self._extract_text_content(pm_response.content)
-                await self.state_sync.on_conference_message(
-                    agent_id="portfolio_manager",
-                    content=pm_content,
-                )
-
-            # Analysts share perspectives
-            for analyst in self.analysts:
-                analyst_prompt = self._build_analyst_discussion_prompt(
+                # PM sets agenda or asks questions
+                pm_prompt = self._build_pm_discussion_prompt(
                    cycle=cycle,
                    tickers=tickers,
                    date=date,
+                    prices=prices,
+                    analyst_results=analyst_results,
+                    risk_assessment=risk_assessment,
                )

-                analyst_msg = Msg(
-                    name="system",
-                    content=analyst_prompt,
-                    role="user",
-                )
-                analyst_response = await analyst.reply(analyst_msg)
+                pm_msg = Msg(name="system", content=pm_prompt, role="user")
+                pm_response = await self.pm.reply(pm_msg)

                if self.state_sync:
-                    analyst_content = self._extract_text_content(
-                        analyst_response.content,
-                    )
+                    pm_content = self._extract_text_content(pm_response.content)
                    await self.state_sync.on_conference_message(
-                        agent_id=analyst.name,
-                        content=analyst_content,
+                        agent_id="portfolio_manager",
+                        content=pm_content,
                    )

-            if self.state_sync:
-                await self.state_sync.on_conference_cycle_end(
-                    cycle=cycle + 1,
-                )
+                # Analysts share perspectives (supports per-round active team updates)
+                for analyst in self._get_active_analysts():
+                    analyst_prompt = self._build_analyst_discussion_prompt(
+                        cycle=cycle,
+                        tickers=tickers,
+                        date=date,
+                    )
+
+                    analyst_msg = Msg(
+                        name="system",
+                        content=analyst_prompt,
+                        role="user",
+                    )
+                    analyst_response = await analyst.reply(analyst_msg)
+
+                    if self.state_sync:
+                        analyst_content = self._extract_text_content(
+                            analyst_response.content,
+                        )
+                        await self.state_sync.on_conference_message(
+                            agent_id=analyst.name,
+                            content=analyst_content,
+                        )
+
+                if self.state_sync:
+                    await self.state_sync.on_conference_cycle_end(
+                        cycle=cycle + 1,
+                    )

        # Generate conference summary by PM
        _log(
@@ -800,6 +935,7 @@ class TradingPipeline:
        self,
        tickers: List[str],
        date: str,
+        active_analysts: Optional[List[Any]] = None,
    ) -> List[Dict[str, Any]]:
        """
        Collect final predictions from all analysts as simple text responses.
@@ -807,14 +943,15 @@ class TradingPipeline:
        """
        _log(
            "Phase 2.2: Analysts generate final structured predictions\n"
-            f"  Starting _collect_final_predictions for {len(self.analysts)} analysts",
+            f"  Starting _collect_final_predictions for {len(active_analysts or self.analysts)} analysts",
        )
        final_predictions = []

-        for i, analyst in enumerate(self.analysts):
+        analysts = active_analysts or self.analysts
+        for i, analyst in enumerate(analysts):
            _log(
                "Phase 2.2: Analysts generate final structured predictions\n"
-                f"  Collecting prediction from analyst {i+1}/{len(self.analysts)}: {analyst.name}",
+                f"  Collecting prediction from analyst {i+1}/{len(analysts)}: {analyst.name}",
            )

            prompt = (
@@ -910,11 +1047,13 @@ class TradingPipeline:
        self,
        tickers: List[str],
        date: str,
+        active_analysts: Optional[List[Any]] = None,
    ) -> List[Dict[str, Any]]:
        """Run all analysts with real-time sync after each completion"""
        results = []
+        analysts = active_analysts or self.analysts

-        for analyst in self.analysts:
+        for analyst in analysts:
            content = (
                f"Analyze the following stocks for date {date}: {', '.join(tickers)}. "
                f"Provide investment signals with confidence scores and reasoning."
@@ -944,15 +1083,107 @@ class TradingPipeline:

        return results

+    async def _run_analysts_parallel(
+        self,
+        tickers: List[str],
+        date: str,
+        active_analysts: Optional[List[Any]] = None,
+    ) -> List[Dict[str, Any]]:
+        """Run all analysts in parallel using TeamCoordinator.
+
+        This method replaces the sequential analyst loop with parallel execution
+        using the TeamCoordinator for orchestration.
+
+        Args:
+            tickers: List of stock tickers to analyze
+            date: Trading date
+            active_analysts: Optional list of analysts to run
+
+        Returns:
+            List of analyst result dictionaries
+        """
+        analysts = active_analysts or self.analysts
+
+        if not analysts:
+            return []
+
+        if not TEAM_COORD_AVAILABLE:
+            _log("TeamCoordinator not available, falling back to sequential execution")
+            return await self._run_analysts_with_sync(
+                tickers=tickers,
+                date=date,
+                active_analysts=active_analysts,
+            )
+
+        _log(
+            f"Phase 1.1: Running {len(analysts)} analysts in parallel "
+            f"[{', '.join(a.name for a in analysts)}]"
+        )
+
+        # Build the analyst prompt
+        content = (
+            f"Analyze the following stocks for date {date}: {', '.join(tickers)}. "
+            f"Provide investment signals with confidence scores and reasoning."
+        )
+
+        # Create coordinator for parallel execution
+        coordinator = TeamCoordinator(
+            participants=analysts,
+            task_content=content,
+        )
+
+        # Run analysts in parallel via TeamCoordinator
+        results = await coordinator.run_phase(
+            "analyst_analysis",
+            metadata={"tickers": tickers, "date": date},
+        )
+
+        # Process results and sync
+        processed_results = []
+        for i, (analyst, result) in enumerate(zip(analysts, results)):
+            if result is not None:
+                extracted = self._extract_result_from_msg(result)
+                processed_results.append(extracted)
+
+                # Sync retrieved memory
+                await self._sync_memory_if_retrieved(analyst)
+
+                # Broadcast agent result via StateSync
+                if self.state_sync:
+                    text_content = self._extract_text_content(result.content)
+                    await self.state_sync.on_agent_complete(
+                        agent_id=analyst.name,
+                        content=text_content,
+                    )
+            else:
+                logger.warning(
+                    "Analyst %s returned no result",
+                    analyst.name,
+                )
+                processed_results.append({
+                    "agent": analyst.name,
+                    "content": "",
+                    "success": False,
+                })
+
+        _log(
+            f"Phase 1.1: Parallel analyst execution complete "
+            f"({len(processed_results)}/{len(analysts)} successful)"
+        )
+
+        return processed_results
+
    async def _run_analysts(
        self,
        tickers: List[str],
        date: str,
+        active_analysts: Optional[List[Any]] = None,
    ) -> List[Dict[str, Any]]:
        """Run all analysts (without sync, for backward compatibility)"""
        results = []
+        analysts = active_analysts or self.analysts

-        for analyst in self.analysts:
+        for analyst in analysts:
            content = (
                f"Analyze the following stocks for date {date}: {', '.join(tickers)}. "
                f"Provide investment signals with confidence scores and reasoning."
@@ -1261,3 +1492,199 @@ class TradingPipeline:
        if decision_texts:
            return "Decisions: " + "; ".join(decision_texts)
        return "Portfolio analysis completed. No trades recommended."
+
+    def load_agents_from_workspace(
+        self,
+        workspace_id: str,
+        agent_factory: Optional[Any] = None,
+    ) -> Dict[str, Any]:
+        """
+        Load agents from workspace using AgentFactory.
+
+        This method supports the new EvoAgent architecture by loading
+        agents from a workspace instead of using hardcoded agents.
+
+        Args:
+            workspace_id: Workspace identifier
+            agent_factory: Optional AgentFactory instance (uses self.agent_factory if None)
+
+        Returns:
+            Dictionary with loaded agents:
+            {
+                "analysts": List[EvoAgent],
+                "risk_manager": EvoAgent,
+                "portfolio_manager": EvoAgent,
+            }
+
+        Raises:
+            ValueError: If workspace doesn't exist or no agents found
+        """
+        factory = agent_factory or self.agent_factory
+        if factory is None:
+            from backend.agents import AgentFactory
+            factory = AgentFactory()
+
+        # Check workspace exists
+        if not factory.workspaces_root.exists():
+            raise ValueError(f"Workspaces root does not exist: {factory.workspaces_root}")
+
+        workspace_dir = factory.workspaces_root / workspace_id
+        if not workspace_dir.exists():
+            raise ValueError(f"Workspace '{workspace_id}' does not exist")
+
+        # Load agents from workspace
+        agents_data = factory.list_agents(workspace_id=workspace_id)
+
+        if not agents_data:
+            raise ValueError(f"No agents found in workspace '{workspace_id}'")
+
+        # Categorize agents by type
+        analysts = []
+        risk_manager = None
+        portfolio_manager = None
+
+        for agent_data in agents_data:
+            agent_type = agent_data.get("agent_type", "unknown")
+            agent_id = agent_data.get("agent_id")
+
+            # Load full agent configuration
+            config_path = Path(agent_data.get("config_path", ""))
+            if config_path.exists():
+                agent = factory.load_agent(agent_id, workspace_id)
+
+                if agent_type.endswith("_analyst"):
+                    analysts.append(agent)
+                elif agent_type == "risk_manager":
+                    risk_manager = agent
+                elif agent_type == "portfolio_manager":
+                    portfolio_manager = agent
+
+        if not analysts:
+            raise ValueError(f"No analysts found in workspace '{workspace_id}'")
+        if risk_manager is None:
+            raise ValueError(f"No risk_manager found in workspace '{workspace_id}'")
+        if portfolio_manager is None:
+            raise ValueError(f"No portfolio_manager found in workspace '{workspace_id}'")
+
+        return {
+            "analysts": analysts,
+            "risk_manager": risk_manager,
+            "portfolio_manager": portfolio_manager,
+        }
+
+    def reload_agents_from_workspace(self, workspace_id: Optional[str] = None) -> None:
+        """
+        Reload all agents from workspace.
+
+        This updates self.analysts, self.risk_manager, and self.pm
+        with agents loaded from the specified workspace.
+
+        Args:
+            workspace_id: Workspace ID (uses self.workspace_id if None)
+        """
+        ws_id = workspace_id or self.workspace_id
+        if not ws_id:
+            raise ValueError("No workspace_id specified")
+
+        loaded = self.load_agents_from_workspace(ws_id)
+
+        self.analysts = loaded["analysts"]
+        self.risk_manager = loaded["risk_manager"]
+        self.pm = loaded["portfolio_manager"]
+        self.workspace_id = ws_id
+
+        logger.info(f"Reloaded {len(self.analysts)} analysts from workspace '{ws_id}'")
+
+    def _runtime_update_status(self, agent: Any, status: str) -> None:
+        if not self.runtime_manager:
+            return
+        agent_id = getattr(agent, "agent_id", None) or getattr(agent, "name", None)
+        if not agent_id:
+            return
+        self.runtime_manager.update_agent_status(agent_id, status, self._session_key)
+
+    def _runtime_batch_status(self, agents: List[Any], status: str) -> None:
+        for agent in agents:
+            self._runtime_update_status(agent, status)
+
+    def _all_analysts(self) -> List[Any]:
+        """Return static analysts plus runtime-created analysts."""
+        return list(self.analysts) + list(self._dynamic_analysts.values())
+
+    def _create_runtime_analyst(self, agent_id: str, analyst_type: str) -> str:
+        """Create one runtime analyst instance."""
+        if analyst_type not in ANALYST_TYPES:
+            return (
+                f"Unknown analyst_type '{analyst_type}'. "
+                f"Available: {', '.join(ANALYST_TYPES.keys())}"
+            )
+        if agent_id in {agent.name for agent in self._all_analysts()}:
+            return f"Analyst '{agent_id}' already exists."
+
+        config_name = getattr(self.pm, "config", {}).get("config_name", "default")
+        project_root = Path(__file__).resolve().parents[2]
+        personas = get_prompt_loader().load_yaml_config("analyst", "personas")
+        persona = personas.get(analyst_type, {})
+        WorkspaceManager(project_root=project_root).ensure_agent_assets(
+            config_name=config_name,
+            agent_id=agent_id,
+            role_seed=persona.get("description", "").strip(),
+            style_seed="\n".join(f"- {item}" for item in persona.get("focus", [])),
+            policy_seed=(
+                "State a clear signal, confidence, and the conditions "
+                "that would invalidate the thesis."
+            ),
+        )
+
+        agent = AnalystAgent(
+            analyst_type=analyst_type,
+            toolkit=create_agent_toolkit(
+                agent_id=agent_id,
+                config_name=config_name,
+                active_skill_dirs=[],
+            ),
+            model=get_agent_model(analyst_type),
+            formatter=get_agent_formatter(analyst_type),
+            agent_id=agent_id,
+            config={"config_name": config_name},
+        )
+        self._dynamic_analysts[agent_id] = agent
+        update_active_analysts(
+            project_root=project_root,
+            config_name=config_name,
+            available_analysts=[item.name for item in self._all_analysts()],
+            add=[agent_id],
+        )
+        return f"Created runtime analyst '{agent_id}' ({analyst_type})."
+
+    def _remove_runtime_analyst(self, agent_id: str) -> str:
+        """Remove one runtime-created analyst instance."""
+        if agent_id not in self._dynamic_analysts:
+            return f"Runtime analyst '{agent_id}' not found."
+        self._dynamic_analysts.pop(agent_id, None)
+        config_name = getattr(self.pm, "config", {}).get("config_name", "default")
+        project_root = Path(__file__).resolve().parents[2]
+        update_active_analysts(
+            project_root=project_root,
+            config_name=config_name,
+            available_analysts=[item.name for item in self._all_analysts()],
+            remove=[agent_id],
+        )
+        return f"Removed runtime analyst '{agent_id}'."
+
+    def _get_active_analysts(self) -> List[Any]:
+        """Resolve active analyst participants from run-scoped team pipeline config."""
+        config_name = getattr(self.pm, "config", {}).get("config_name", "default")
+        project_root = Path(__file__).resolve().parents[2]
+        analyst_map = {agent.name: agent for agent in self._all_analysts()}
+        active_ids = resolve_active_analysts(
+            project_root=project_root,
+            config_name=config_name,
+            available_analysts=list(analyst_map.keys()),
+        )
+        return [analyst_map[agent_id] for agent_id in active_ids if agent_id in analyst_map]
+
+    def _runtime_log_event(self, event: str, details: Optional[Dict[str, Any]] = None) -> None:
+        if not self.runtime_manager:
+            return
+        self.runtime_manager.log_event(event, details)
--- a/backend/core/pipeline_runner.py
+++ b/backend/core/pipeline_runner.py
@@ -0,0 +1,489 @@
+# -*- coding: utf-8 -*-
+"""
+Pipeline Runner - Independent trading pipeline execution
+
+This module provides functions to start/stop trading pipelines
+that can be called from the REST API.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import os
+from contextlib import AsyncExitStack
+from pathlib import Path
+from typing import Any, Dict, Optional, Callable
+
+from backend.agents import AnalystAgent, PMAgent, RiskAgent
+from backend.agents.skills_manager import SkillsManager
+from backend.agents.toolkit_factory import create_agent_toolkit, load_agent_profiles
+from backend.agents.prompt_loader import get_prompt_loader
+from backend.agents.workspace_manager import WorkspaceManager
+from backend.config.constants import ANALYST_TYPES
+from backend.core.pipeline import TradingPipeline
+from backend.core.scheduler import BacktestScheduler, Scheduler
+from backend.llm.models import get_agent_formatter, get_agent_model
+from backend.runtime.manager import (
+    TradingRuntimeManager,
+    set_global_runtime_manager,
+    clear_global_runtime_manager,
+    set_shutdown_event,
+    clear_shutdown_event,
+    is_shutdown_requested,
+)
+from backend.services.market import MarketService
+from backend.services.storage import StorageService
+from backend.services.gateway import Gateway
+from backend.utils.settlement import SettlementCoordinator
+
+_prompt_loader = get_prompt_loader()
+
+# Global gateway reference for cleanup
+_gateway_instance: Optional[Gateway] = None
+
+
+def _set_gateway(gateway: Optional[Gateway]) -> None:
+    """Set global gateway reference."""
+    global _gateway_instance
+    _gateway_instance = gateway
+
+
+def stop_gateway() -> None:
+    """Stop the running gateway if exists."""
+    global _gateway_instance
+    if _gateway_instance is not None:
+        try:
+            _gateway_instance.stop()
+        except Exception as e:
+            import logging
+            logging.getLogger(__name__).error(f"Error stopping gateway: {e}")
+        finally:
+            _gateway_instance = None
+
+
+def create_long_term_memory(agent_name: str, run_id: str, run_dir: Path):
+    """Create ReMeTaskLongTermMemory for an agent."""
+    try:
+        from agentscope.memory import ReMeTaskLongTermMemory
+        from agentscope.model import DashScopeChatModel
+        from agentscope.embedding import DashScopeTextEmbedding
+    except ImportError:
+        return None
+
+    api_key = os.getenv("MEMORY_API_KEY")
+    if not api_key:
+        return None
+
+    memory_dir = str(run_dir / "memory")
+
+    return ReMeTaskLongTermMemory(
+        agent_name=agent_name,
+        user_name=agent_name,
+        model=DashScopeChatModel(
+            model_name=os.getenv("MEMORY_MODEL_NAME", "qwen3-max"),
+            api_key=api_key,
+            stream=False,
+        ),
+        embedding_model=DashScopeTextEmbedding(
+            model_name=os.getenv("MEMORY_EMBEDDING_MODEL", "text-embedding-v4"),
+            api_key=api_key,
+            dimensions=1024,
+        ),
+        **{
+            "vector_store.default.backend": "local",
+            "vector_store.default.params.store_dir": memory_dir,
+        },
+    )
+
+
+def create_agents(
+    run_id: str,
+    run_dir: Path,
+    initial_cash: float,
+    margin_requirement: float,
+    enable_long_term_memory: bool = False,
+):
+    """Create all agents for the system."""
+    analysts = []
+    long_term_memories = []
+
+    # Initialize workspace manager and assets
+    workspace_manager = WorkspaceManager()
+    workspace_manager.initialize_default_assets(
+        config_name=run_id,
+        agent_ids=list(ANALYST_TYPES.keys()) + ["risk_manager", "portfolio_manager"],
+        analyst_personas=_prompt_loader.load_yaml_config("analyst", "personas"),
+    )
+
+    profiles = load_agent_profiles()
+    skills_manager = SkillsManager()
+    active_skill_map = skills_manager.prepare_active_skills(
+        config_name=run_id,
+        agent_defaults={
+            agent_id: profile.get("skills", [])
+            for agent_id, profile in profiles.items()
+        },
+    )
+
+    # Create analyst agents
+    for analyst_type in ANALYST_TYPES:
+        model = get_agent_model(analyst_type)
+        formatter = get_agent_formatter(analyst_type)
+        toolkit = create_agent_toolkit(
+            analyst_type,
+            run_id,
+            active_skill_dirs=active_skill_map.get(analyst_type, []),
+        )
+
+        long_term_memory = None
+        if enable_long_term_memory:
+            long_term_memory = create_long_term_memory(analyst_type, run_id, run_dir)
+            if long_term_memory:
+                long_term_memories.append(long_term_memory)
+
+        analyst = AnalystAgent(
+            analyst_type=analyst_type,
+            toolkit=toolkit,
+            model=model,
+            formatter=formatter,
+            agent_id=analyst_type,
+            config={"config_name": run_id},
+            long_term_memory=long_term_memory,
+        )
+        analysts.append(analyst)
+
+    # Create risk manager
+    risk_long_term_memory = None
+    if enable_long_term_memory:
+        risk_long_term_memory = create_long_term_memory("risk_manager", run_id, run_dir)
+        if risk_long_term_memory:
+            long_term_memories.append(risk_long_term_memory)
+
+    risk_manager = RiskAgent(
+        model=get_agent_model("risk_manager"),
+        formatter=get_agent_formatter("risk_manager"),
+        name="risk_manager",
+        config={"config_name": run_id},
+        long_term_memory=risk_long_term_memory,
+        toolkit=create_agent_toolkit(
+            "risk_manager",
+            run_id,
+            active_skill_dirs=active_skill_map.get("risk_manager", []),
+        ),
+    )
+
+    # Create portfolio manager
+    pm_long_term_memory = None
+    if enable_long_term_memory:
+        pm_long_term_memory = create_long_term_memory("portfolio_manager", run_id, run_dir)
+        if pm_long_term_memory:
+            long_term_memories.append(pm_long_term_memory)
+
+    portfolio_manager = PMAgent(
+        name="portfolio_manager",
+        model=get_agent_model("portfolio_manager"),
+        formatter=get_agent_formatter("portfolio_manager"),
+        initial_cash=initial_cash,
+        margin_requirement=margin_requirement,
+        config={"config_name": run_id},
+        long_term_memory=pm_long_term_memory,
+        toolkit_factory=create_agent_toolkit,
+        toolkit_factory_kwargs={
+            "active_skill_dirs": active_skill_map.get("portfolio_manager", []),
+        },
+    )
+
+    return analysts, risk_manager, portfolio_manager, long_term_memories
+
+
+async def run_pipeline(
+    run_id: str,
+    run_dir: Path,
+    bootstrap: Dict[str, Any],
+    stop_event: asyncio.Event,
+    message_callback: Optional[Callable[[str, Any], None]] = None,
+) -> None:
+    """
+    Run the trading pipeline with the given configuration.
+
+    Service Startup Order:
+        Phase 1: WebSocket Server - Frontend can connect
+        Phase 2: Market Service - Price data starts flowing
+        Phase 3: Agent Runtime - Create all agents
+        Phase 4: Pipeline & Scheduler - Trading logic ready
+        Phase 5: Gateway Fully Operational - All systems running
+
+    Args:
+        run_id: Unique run identifier (timestamp)
+        run_dir: Run directory path
+        bootstrap: Bootstrap configuration
+        stop_event: Event to signal pipeline stop
+        message_callback: Optional callback for sending messages to clients
+    """
+    import logging
+    logger = logging.getLogger(__name__)
+
+    # Set global shutdown event
+    set_shutdown_event(stop_event)
+
+    logger.info(f"[Pipeline {run_id}] ======================================")
+    logger.info(f"[Pipeline {run_id}] Starting with 5-phase initialization...")
+    logger.info(f"[Pipeline {run_id}] ======================================")
+
+    try:
+        # Extract config values
+        tickers = bootstrap.get("tickers", ["AAPL", "MSFT"])
+        initial_cash = float(bootstrap.get("initial_cash", 100000.0))
+        margin_requirement = float(bootstrap.get("margin_requirement", 0.0))
+        max_comm_cycles = int(bootstrap.get("max_comm_cycles", 2))
+        schedule_mode = bootstrap.get("schedule_mode", "daily")
+        trigger_time = bootstrap.get("trigger_time", "09:30")
+        interval_minutes = int(bootstrap.get("interval_minutes", 60))
+        heartbeat_interval = int(bootstrap.get("heartbeat_interval", 0))
+        mode = bootstrap.get("mode", "live")
+        start_date = bootstrap.get("start_date")
+        end_date = bootstrap.get("end_date")
+        enable_memory = bootstrap.get("enable_memory", False)
+        enable_mock = bootstrap.get("enable_mock", False)
+
+        is_backtest = mode == "backtest"
+        is_mock = enable_mock or mode == "mock" or (not is_backtest and os.getenv("MOCK_MODE", "false").lower() == "true")
+
+        # ======================================================================
+        # PHASE 0: Initialize runtime manager
+        # ======================================================================
+        logger.info("[Phase 0/5] Initializing runtime manager...")
+
+        from backend.api.runtime import runtime_manager
+
+        if runtime_manager is None:
+            runtime_manager = TradingRuntimeManager(
+                config_name=run_id,
+                run_dir=run_dir,
+                bootstrap=bootstrap,
+            )
+            runtime_manager.prepare_run()
+
+        set_global_runtime_manager(runtime_manager)
+
+        # Register runtime manager with API
+        from backend.api.runtime import register_runtime_manager
+        register_runtime_manager(runtime_manager)
+
+        # ======================================================================
+        # PHASE 1 & 2: Create infrastructure services (Market, Storage)
+        # These will be started by Gateway in the correct order
+        # ======================================================================
+        logger.info("[Phase 1-2/5] Creating infrastructure services...")
+
+        # Create storage service
+        storage_service = StorageService(
+            dashboard_dir=run_dir / "team_dashboard",
+            initial_cash=initial_cash,
+            config_name=run_id,
+        )
+
+        if not storage_service.files["summary"].exists():
+            storage_service.initialize_empty_dashboard()
+        else:
+            storage_service.update_leaderboard_model_info()
+
+        # Create market service (data source)
+        market_service = MarketService(
+            tickers=tickers,
+            poll_interval=10,
+            mock_mode=is_mock and not is_backtest,
+            backtest_mode=is_backtest,
+            api_key=os.getenv("FINNHUB_API_KEY") if not is_mock and not is_backtest else None,
+            backtest_start_date=start_date if is_backtest else None,
+            backtest_end_date=end_date if is_backtest else None,
+        )
+
+        # ======================================================================
+        # PHASE 3: Create Agent Runtime
+        # ======================================================================
+        logger.info("[Phase 3/5] Creating agent runtime...")
+
+        analysts, risk_manager, pm, long_term_memories = create_agents(
+            run_id=run_id,
+            run_dir=run_dir,
+            initial_cash=initial_cash,
+            margin_requirement=margin_requirement,
+            enable_long_term_memory=enable_memory,
+        )
+
+        # Register agents with runtime manager
+        for agent in analysts + [risk_manager, pm]:
+            agent_id = getattr(agent, "agent_id", None) or getattr(agent, "name", None)
+            if agent_id:
+                runtime_manager.register_agent(agent_id)
+
+        # Load portfolio state
+        portfolio_state = storage_service.load_portfolio_state()
+        pm.load_portfolio_state(portfolio_state)
+
+        # Create settlement coordinator
+        settlement_coordinator = SettlementCoordinator(
+            storage=storage_service,
+            initial_capital=initial_cash,
+        )
+
+        # ======================================================================
+        # PHASE 4: Create Pipeline & Scheduler
+        # ======================================================================
+        logger.info("[Phase 4/5] Creating pipeline and scheduler...")
+
+        # Create pipeline
+        pipeline = TradingPipeline(
+            analysts=analysts,
+            risk_manager=risk_manager,
+            portfolio_manager=pm,
+            settlement_coordinator=settlement_coordinator,
+            max_comm_cycles=max_comm_cycles,
+            runtime_manager=runtime_manager,
+        )
+
+        # Create scheduler
+        scheduler_callback = None
+        live_scheduler = None
+
+        if is_backtest:
+            backtest_scheduler = BacktestScheduler(
+                start_date=start_date,
+                end_date=end_date,
+                trading_calendar="NYSE",
+                delay_between_days=0.5,
+            )
+            trading_dates = backtest_scheduler.get_trading_dates()
+
+            async def scheduler_callback_fn(callback):
+                await backtest_scheduler.start(callback)
+
+            scheduler_callback = scheduler_callback_fn
+        else:
+            # Live mode
+            live_scheduler = Scheduler(
+                mode=schedule_mode,
+                trigger_time=trigger_time,
+                interval_minutes=interval_minutes,
+                heartbeat_interval=heartbeat_interval if heartbeat_interval > 0 else None,
+                config={"config_name": run_id},
+            )
+
+            async def scheduler_callback_fn(callback):
+                await live_scheduler.start(callback)
+
+            scheduler_callback = scheduler_callback_fn
+
+        # ======================================================================
+        # PHASE 5: Start Gateway (WebSocket → Market → Scheduler)
+        # Gateway.start() will handle the final startup sequence:
+        #   - WebSocket Server first (frontend can connect)
+        #   - Market Service second (price data flows)
+        #   - Scheduler last (trading begins)
+        # ======================================================================
+        logger.info("[Phase 5/5] Starting Gateway (WebSocket → Market → Scheduler)...")
+
+        gateway = Gateway(
+            market_service=market_service,
+            storage_service=storage_service,
+            pipeline=pipeline,
+            scheduler_callback=scheduler_callback,
+            config={
+                "mode": mode,
+                "mock_mode": is_mock,
+                "backtest_mode": is_backtest,
+                "tickers": tickers,
+                "config_name": run_id,
+                "schedule_mode": schedule_mode,
+                "interval_minutes": interval_minutes,
+                "trigger_time": trigger_time,
+                "heartbeat_interval": heartbeat_interval,
+                "initial_cash": initial_cash,
+                "margin_requirement": margin_requirement,
+                "max_comm_cycles": max_comm_cycles,
+                "enable_memory": enable_memory,
+            },
+            scheduler=live_scheduler,
+        )
+        _set_gateway(gateway)
+
+        # Start pipeline execution
+        async with AsyncExitStack() as stack:
+            # Enter long-term memory contexts
+            for memory in long_term_memories:
+                await stack.enter_async_context(memory)
+
+            # Start Gateway - this will execute the 4-phase startup:
+            # Phase 1: WebSocket Server (frontend can connect immediately)
+            # Phase 2: Market Service (price updates start flowing)
+            # Phase 3: Market Status Monitor
+            # Phase 4: Scheduler (trading cycles begin)
+            gateway_task = asyncio.create_task(
+                gateway.start(host="0.0.0.0", port=8765)
+            )
+            logger.info("[Pipeline] Gateway startup initiated on ws://localhost:8765")
+
+            # Wait for Gateway to fully initialize all phases
+            await asyncio.sleep(0.5)
+
+            # Define the trading cycle callback
+            async def trading_cycle(session_key: str) -> None:
+                """Execute one trading cycle."""
+                if is_shutdown_requested():
+                    return
+
+                runtime_manager.set_session_key(session_key)
+                runtime_manager.log_event("cycle:start", {"session": session_key})
+
+                try:
+                    # Fetch market data
+                    market_data = await market_service.get_all_data()
+
+                    # Run pipeline
+                    await pipeline.run_cycle(
+                        session_key=session_key,
+                        market_data=market_data,
+                    )
+
+                    runtime_manager.log_event("cycle:complete", {"session": session_key})
+
+                except Exception as e:
+                    runtime_manager.log_event("cycle:error", {"error": str(e)})
+                    raise
+
+            # Start scheduler
+            if scheduler_callback:
+                await scheduler_callback(trading_cycle)
+
+            # Wait for stop signal
+            while not stop_event.is_set():
+                await asyncio.sleep(1)
+
+            # Cancel gateway task
+            if not gateway_task.done():
+                gateway_task.cancel()
+                try:
+                    await gateway_task
+                except asyncio.CancelledError:
+                    pass
+
+    except asyncio.CancelledError:
+        # Handle cancellation gracefully
+        raise
+    finally:
+        # Cleanup
+        logger.info("[Pipeline] Cleaning up...")
+
+        # Stop Gateway
+        try:
+            stop_gateway()
+            logger.info("[Pipeline] Gateway stopped")
+        except Exception as e:
+            logger.error(f"[Pipeline] Error stopping gateway: {e}")
+
+        clear_shutdown_event()
+        clear_global_runtime_manager()
+        from backend.api.runtime import unregister_runtime_manager
+        unregister_runtime_manager()
+        logger.info("[Pipeline] Cleanup complete")
--- a/backend/core/scheduler.py
+++ b/backend/core/scheduler.py
@@ -4,7 +4,7 @@ Scheduler - Market-aware trigger system for trading cycles
 """
 import asyncio
 import logging
-from datetime import datetime, timedelta
+from datetime import datetime, time, timedelta
 from typing import Any, Callable, Optional
 from zoneinfo import ZoneInfo

@@ -28,16 +28,21 @@ class Scheduler:
        mode: str = "daily",
        trigger_time: Optional[str] = None,
        interval_minutes: Optional[int] = None,
+        heartbeat_interval: Optional[int] = None,
        config: Optional[dict] = None,
    ):
        self.mode = mode
        self.trigger_time = trigger_time or "09:30"  # NYSE timezone
        self.trigger_now = self.trigger_time == "now"
        self.interval_minutes = interval_minutes or 60
+        self.heartbeat_interval = heartbeat_interval  # e.g. 3600 = 1 hour
        self.config = config or {}

        self.running = False
        self._task: Optional[asyncio.Task] = None
+        self._heartbeat_task: Optional[asyncio.Task] = None
+        self._callback: Optional[Callable] = None
+        self._heartbeat_callback: Optional[Callable] = None

    def _now_nyse(self) -> datetime:
        """Get current time in NYSE timezone"""
@@ -52,6 +57,15 @@ class Scheduler:
        )
        return len(valid_days) > 0

+    def _is_trading_hours(self, now: datetime) -> bool:
+        """Check if current time is within NYSE trading hours (9:30-16:00 ET)."""
+        market_time = now.time()
+        return time(9, 30) <= market_time <= time(16, 0)
+
+    def set_heartbeat_callback(self, callback: Callable) -> None:
+        """Register callback for heartbeat triggers."""
+        self._heartbeat_callback = callback
+
    def _next_trading_day(self, from_date: datetime) -> datetime:
        """Find the next trading day from given date"""
        check_date = from_date
@@ -68,18 +82,100 @@ class Scheduler:
            return

        self.running = True
+        self._callback = callback
+        self._schedule_task()

-        if self.mode == "daily":
-            self._task = asyncio.create_task(self._run_daily(callback))
-        elif self.mode == "intraday":
-            self._task = asyncio.create_task(self._run_intraday(callback))
-        else:
-            raise ValueError(f"Unknown scheduler mode: {self.mode}")
+        # Start heartbeat loop if configured
+        if self.heartbeat_interval and self._heartbeat_callback:
+            self._heartbeat_task = asyncio.create_task(self._run_heartbeat_loop())
+            logger.info(
+                f"Heartbeat loop started: interval={self.heartbeat_interval}s",
+            )

        logger.info(
            f"Scheduler started: mode={self.mode}, timezone=America/New_York",
        )

+    def _schedule_task(self):
+        """Create the active scheduler task for the current mode."""
+        if not self._callback:
+            raise ValueError("Scheduler callback is not set")
+
+        if self._task:
+            self._task.cancel()
+            self._task = None
+
+        if self.mode == "daily":
+            self._task = asyncio.create_task(self._run_daily(self._callback))
+        elif self.mode == "intraday":
+            self._task = asyncio.create_task(
+                self._run_intraday(self._callback),
+            )
+        else:
+            raise ValueError(f"Unknown scheduler mode: {self.mode}")
+
+    def reconfigure(
+        self,
+        *,
+        mode: Optional[str] = None,
+        trigger_time: Optional[str] = None,
+        interval_minutes: Optional[int] = None,
+    ) -> bool:
+        """Update scheduler parameters in-place and restart its timing loop."""
+        changed = False
+
+        if mode and mode != self.mode:
+            self.mode = mode
+            changed = True
+
+        if trigger_time and trigger_time != self.trigger_time:
+            self.trigger_time = trigger_time
+            self.trigger_now = self.trigger_time == "now"
+            changed = True
+
+        if (
+            interval_minutes is not None
+            and interval_minutes > 0
+            and interval_minutes != self.interval_minutes
+        ):
+            self.interval_minutes = interval_minutes
+            changed = True
+
+        if changed and self.running and self._callback:
+            self._schedule_task()
+            logger.info(
+                "Scheduler reconfigured: mode=%s, trigger_time=%s, interval_minutes=%s",
+                self.mode,
+                self.trigger_time,
+                self.interval_minutes,
+            )
+
+        return changed
+
+    async def _run_heartbeat_loop(self):
+        """Run heartbeat checks on a separate interval during trading hours."""
+        while self.running:
+            now = self._now_nyse()
+            if self._is_trading_day(now) and self._is_trading_hours(now):
+                if self._heartbeat_callback:
+                    try:
+                        current_date = now.strftime("%Y-%m-%d")
+                        logger.debug(
+                            f"[Heartbeat] Triggering heartbeat check for {current_date}",
+                        )
+                        await self._heartbeat_callback(date=current_date)
+                    except Exception as e:
+                        logger.error(
+                            f"[Heartbeat] Callback failed: {e}",
+                            exc_info=True,
+                        )
+                else:
+                    logger.warning(
+                        "[Heartbeat] Callback not set, skipping heartbeat",
+                    )
+
+            await asyncio.sleep(self.heartbeat_interval)
+
    async def _run_daily(self, callback: Callable):
        """Run once per trading day at specified time (NYSE timezone)"""
        first_run = True
@@ -154,6 +250,9 @@ class Scheduler:
        if self._task:
            self._task.cancel()
            self._task = None
+        if self._heartbeat_task:
+            self._heartbeat_task.cancel()
+            self._heartbeat_task = None
        logger.info("Scheduler stopped")


--- a/backend/core/state_sync.py
+++ b/backend/core/state_sync.py
@@ -47,6 +47,10 @@ class StateSync:
        """Set current simulation date for backtest-compatible timestamps"""
        self._simulation_date = date

+    def clear_simulation_date(self):
+        """Disable backtest timestamp simulation and use wall-clock time."""
+        self._simulation_date = None
+
    def _get_timestamp_ms(self) -> int:
        """
        Get timestamp in milliseconds.
@@ -97,9 +101,21 @@ class StateSync:
        if not self._enabled:
            return

-        # Ensure timestamp exists (use simulation date if in backtest mode)
+        # Ensure timestamp exists. Prefer explicit millisecond timestamps so
+        # frontend displays local wall time correctly instead of date-only UTC.
        if "timestamp" not in event:
-            if self._simulation_date:
+            ts_ms = event.get("ts")
+            if ts_ms is not None:
+                try:
+                    event["timestamp"] = datetime.fromtimestamp(
+                        float(ts_ms) / 1000.0,
+                    ).isoformat()
+                except (TypeError, ValueError, OSError):
+                    if self._simulation_date:
+                        event["timestamp"] = f"{self._simulation_date}"
+                    else:
+                        event["timestamp"] = datetime.now().isoformat()
+            elif self._simulation_date:
                event["timestamp"] = f"{self._simulation_date}"
            else:
                event["timestamp"] = datetime.now().isoformat()
@@ -238,9 +254,12 @@ class StateSync:
        """Called at start of trading cycle"""
        self._state["current_date"] = date
        self._state["status"] = "running"
-        self.set_simulation_date(
-            date,
-        )  # Set for backtest-compatible timestamps
+        if self._state.get("server_mode") == "backtest":
+            self.set_simulation_date(
+                date,
+            )  # Set for backtest-compatible timestamps
+        else:
+            self.clear_simulation_date()

        await self.emit(
            {
@@ -411,7 +430,9 @@ class StateSync:

        Useful for: frontend reconnection or restoring from saved state
        """
-        feed_history = self._state.get("feed_history", [])
+        feed_history = self.storage.runtime_db.get_recent_feed_events(
+            limit=self.storage.max_feed_history,
+        ) or self._state.get("feed_history", [])

        # feed_history is newest-first, need to reverse for chronological replay # noqa: E501
        for event in reversed(feed_history):
@@ -434,11 +455,22 @@ class StateSync:
        Returns:
            Dictionary suitable for sending to frontend
        """
+        feed_history = self.storage.runtime_db.get_recent_feed_events(
+            limit=self.storage.max_feed_history,
+        ) or self._state.get("feed_history", [])
+        last_day_history = self.storage.runtime_db.get_last_day_feed_events(
+            current_date=self._state.get("current_date"),
+            limit=self.storage.max_feed_history,
+        ) or self._state.get("last_day_history", [])
+
        payload = {
            "server_mode": self._state.get("server_mode", "live"),
            "is_mock_mode": self._state.get("is_mock_mode", False),
            "is_backtest": self._state.get("is_backtest", False),
-            "feed_history": self._state.get("feed_history", []),
+            "tickers": self._state.get("tickers"),
+            "runtime_config": self._state.get("runtime_config"),
+            "feed_history": feed_history,
+            "last_day_history": last_day_history,
            "current_date": self._state.get("current_date"),
            "trading_days_total": self._state.get("trading_days_total", 0),
            "trading_days_completed": self._state.get(
@@ -452,6 +484,7 @@ class StateSync:
            "portfolio": self._state.get("portfolio", {}),
            "realtime_prices": self._state.get("realtime_prices", {}),
            "data_sources": self._state.get("data_sources", {}),
+            "price_history": self._state.get("price_history", {}),
        }

        if include_dashboard:
--- a/backend/data/historical_price_manager.py
+++ b/backend/data/historical_price_manager.py
@@ -7,6 +7,7 @@ from datetime import datetime
 from typing import Callable, Dict, List, Optional

 import pandas as pd
+from backend.data.market_store import MarketStore
 from backend.data.provider_utils import normalize_symbol
 from backend.data.provider_router import get_provider_router

@@ -26,6 +27,7 @@ class HistoricalPriceManager:
        self.close_prices = {}
        self.running = False
        self._router = get_provider_router()
+        self._market_store = MarketStore()

    def subscribe(
        self,
@@ -58,21 +60,48 @@ class HistoricalPriceManager:
            logger.warning(f"Failed to load CSV for {symbol}: {e}")
            return None

+    def _load_from_market_db(
+        self,
+        symbol: str,
+        start_date: str,
+        end_date: str,
+    ) -> Optional[pd.DataFrame]:
+        """Load price data from the long-lived market research database."""
+        try:
+            rows = self._market_store.get_ohlc(symbol, start_date, end_date)
+            if not rows:
+                return None
+            df = pd.DataFrame(rows)
+            if df.empty or "date" not in df.columns:
+                return None
+            df["Date"] = pd.to_datetime(df["date"])
+            df.set_index("Date", inplace=True)
+            df.sort_index(inplace=True)
+            return df
+        except Exception as e:
+            logger.warning(f"Failed to load market DB data for {symbol}: {e}")
+            return None
+
    def preload_data(self, start_date: str, end_date: str):
-        """Preload historical data from local CSV files."""
+        """Preload historical data from market DB first, then local CSV."""
        logger.info(f"Preloading data: {start_date} to {end_date}")

        for symbol in self.subscribed_symbols:
            if symbol in self._price_cache:
                continue

-            # Load from local CSV file directly
+            df = self._load_from_market_db(symbol, start_date, end_date)
+            if df is not None and not df.empty:
+                self._price_cache[symbol] = df
+                logger.info(f"Loaded {symbol} from market DB: {len(df)} records")
+                continue
+
            df = self._load_from_csv(symbol)
            if df is not None and not df.empty:
                self._price_cache[symbol] = df
                logger.info(f"Loaded {symbol} from CSV: {len(df)} records")
            else:
-                logger.warning(f"No CSV data for {symbol}")
+                logger.warning(f"No market DB or CSV data for {symbol}")

    def set_date(self, date: str):
        """Set current trading date and update prices"""
--- a/backend/data/market_ingest.py
+++ b/backend/data/market_ingest.py
@@ -0,0 +1,149 @@
+# -*- coding: utf-8 -*-
+"""Ingest Polygon market data into the long-lived research warehouse."""
+
+from __future__ import annotations
+
+from datetime import datetime, timedelta, timezone
+from typing import Iterable
+
+from backend.data.market_store import MarketStore
+from backend.data.news_alignment import align_news_for_symbol
+from backend.data.polygon_client import (
+    fetch_news,
+    fetch_ohlc,
+    fetch_ticker_details,
+)
+from backend.data.provider_utils import normalize_symbol
+
+
+def _today_utc() -> str:
+    return datetime.now(timezone.utc).date().isoformat()
+
+
+def _default_start(years: int = 2) -> str:
+    return (datetime.now(timezone.utc).date() - timedelta(days=years * 366)).isoformat()
+
+
+def ingest_ticker_history(
+    symbol: str,
+    *,
+    start_date: str | None = None,
+    end_date: str | None = None,
+    store: MarketStore | None = None,
+) -> dict:
+    """Fetch and persist Polygon OHLC + news for a ticker."""
+    ticker = normalize_symbol(symbol)
+    start = start_date or _default_start()
+    end = end_date or _today_utc()
+    market_store = store or MarketStore()
+
+    details = fetch_ticker_details(ticker)
+    market_store.upsert_ticker(
+        symbol=ticker,
+        name=details.get("name"),
+        sector=details.get("sic_description"),
+        is_active=bool(details.get("active", True)),
+    )
+
+    ohlc_rows = fetch_ohlc(ticker, start, end)
+    news_rows = fetch_news(ticker, start, end)
+    price_count = market_store.upsert_ohlc(ticker, ohlc_rows, source="polygon")
+    news_count = market_store.upsert_news(ticker, news_rows, source="polygon")
+    aligned_count = align_news_for_symbol(market_store, ticker)
+    market_store.update_fetch_watermark(symbol=ticker, price_date=end, news_date=end)
+
+    return {
+        "symbol": ticker,
+        "start_date": start,
+        "end_date": end,
+        "prices": price_count,
+        "news": news_count,
+        "aligned": aligned_count,
+    }
+
+
+def update_ticker_incremental(
+    symbol: str,
+    *,
+    end_date: str | None = None,
+    store: MarketStore | None = None,
+) -> dict:
+    """Incrementally fetch OHLC + news since the last watermark."""
+    ticker = normalize_symbol(symbol)
+    market_store = store or MarketStore()
+    watermarks = market_store.get_ticker_watermarks(ticker)
+    end = end_date or _today_utc()
+    start_prices = (
+        (datetime.fromisoformat(watermarks["last_price_fetch"]) + timedelta(days=1)).date().isoformat()
+        if watermarks.get("last_price_fetch")
+        else _default_start()
+    )
+    start_news = (
+        (datetime.fromisoformat(watermarks["last_news_fetch"]) + timedelta(days=1)).date().isoformat()
+        if watermarks.get("last_news_fetch")
+        else _default_start()
+    )
+
+    details = fetch_ticker_details(ticker)
+    market_store.upsert_ticker(
+        symbol=ticker,
+        name=details.get("name"),
+        sector=details.get("sic_description"),
+        is_active=bool(details.get("active", True)),
+    )
+
+    ohlc_rows = [] if start_prices > end else fetch_ohlc(ticker, start_prices, end)
+    news_rows = [] if start_news > end else fetch_news(ticker, start_news, end)
+    price_count = market_store.upsert_ohlc(ticker, ohlc_rows, source="polygon") if ohlc_rows else 0
+    news_count = market_store.upsert_news(ticker, news_rows, source="polygon") if news_rows else 0
+    aligned_count = align_news_for_symbol(market_store, ticker)
+    market_store.update_fetch_watermark(
+        symbol=ticker,
+        price_date=end if ohlc_rows or watermarks.get("last_price_fetch") else None,
+        news_date=end if news_rows or watermarks.get("last_news_fetch") else None,
+    )
+
+    return {
+        "symbol": ticker,
+        "start_price_date": start_prices,
+        "start_news_date": start_news,
+        "end_date": end,
+        "prices": price_count,
+        "news": news_count,
+        "aligned": aligned_count,
+    }
+
+
+def ingest_symbols(
+    symbols: Iterable[str],
+    *,
+    mode: str = "incremental",
+    start_date: str | None = None,
+    end_date: str | None = None,
+    store: MarketStore | None = None,
+) -> list[dict]:
+    """Fetch Polygon data for a list of tickers."""
+    market_store = store or MarketStore()
+    results = []
+    for symbol in symbols:
+        ticker = normalize_symbol(symbol)
+        if not ticker:
+            continue
+        if mode == "full":
+            results.append(
+                ingest_ticker_history(
+                    ticker,
+                    start_date=start_date,
+                    end_date=end_date,
+                    store=market_store,
+                )
+            )
+        else:
+            results.append(
+                update_ticker_incremental(
+                    ticker,
+                    end_date=end_date,
+                    store=market_store,
+                )
+            )
+    return results
--- a/backend/data/market_store.py
+++ b/backend/data/market_store.py
--- a/backend/data/news_alignment.py
+++ b/backend/data/news_alignment.py
@@ -0,0 +1,64 @@
+# -*- coding: utf-8 -*-
+"""Align persisted news to the nearest NYSE trading date."""
+
+from __future__ import annotations
+
+from datetime import time
+
+import pandas as pd
+import pandas_market_calendars as mcal
+
+from backend.data.market_store import MarketStore
+
+
+NYSE_CALENDAR = mcal.get_calendar("NYSE")
+
+
+def _next_trading_day(date_str: str) -> str:
+    start = pd.Timestamp(date_str).tz_localize(None)
+    sessions = NYSE_CALENDAR.valid_days(
+        start_date=(start - pd.Timedelta(days=1)).strftime("%Y-%m-%d"),
+        end_date=(start + pd.Timedelta(days=10)).strftime("%Y-%m-%d"),
+    )
+    future = [
+        pd.Timestamp(day).tz_localize(None).strftime("%Y-%m-%d")
+        for day in sessions
+        if pd.Timestamp(day).tz_localize(None) >= start
+    ]
+    return future[0] if future else date_str
+
+
+def resolve_trade_date(published_utc: str | None) -> str | None:
+    """Map a published timestamp to an NYSE trade date."""
+    if not published_utc:
+        return None
+    timestamp = pd.to_datetime(published_utc, utc=True, errors="coerce")
+    if pd.isna(timestamp):
+        return None
+    nyse_time = timestamp.tz_convert("America/New_York")
+    candidate = nyse_time.date().isoformat()
+    valid_days = NYSE_CALENDAR.valid_days(start_date=candidate, end_date=candidate)
+    if len(valid_days) == 0:
+        return _next_trading_day(candidate)
+    if nyse_time.time() >= time(16, 0):
+        return _next_trading_day((nyse_time + pd.Timedelta(days=1)).date().isoformat())
+    return candidate
+
+
+def align_news_for_symbol(store: MarketStore, symbol: str, *, limit: int = 5000) -> int:
+    """Fill missing trade_date values for one ticker."""
+    pending = store.get_news_without_trade_date(symbol, limit=limit)
+    updates = []
+    for row in pending:
+        trade_date = resolve_trade_date(row.get("published_utc"))
+        if trade_date:
+            updates.append(
+                {
+                    "news_id": row["news_id"],
+                    "symbol": row["symbol"],
+                    "trade_date": trade_date,
+                }
+            )
+    if not updates:
+        return 0
+    return store.set_trade_dates(updates)
--- a/backend/data/polygon_client.py
+++ b/backend/data/polygon_client.py
@@ -0,0 +1,161 @@
+# -*- coding: utf-8 -*-
+"""Polygon client used for long-lived market research ingestion."""
+
+from __future__ import annotations
+
+import os
+import time
+from datetime import datetime, timezone
+from typing import Any, Optional
+
+import requests
+
+
+BASE = "https://api.polygon.io"
+
+
+def _headers() -> dict[str, str]:
+    api_key = os.getenv("POLYGON_API_KEY", "").strip()
+    if not api_key:
+        raise ValueError("Missing required API key: POLYGON_API_KEY")
+    return {"Authorization": f"Bearer {api_key}"}
+
+
+def http_get(
+    url: str,
+    params: Optional[dict[str, Any]] = None,
+    *,
+    max_retries: int = 8,
+    backoff: float = 2.0,
+) -> requests.Response:
+    """HTTP GET with exponential backoff and 429 handling."""
+    for attempt in range(max_retries):
+        try:
+            response = requests.get(
+                url,
+                params=params or {},
+                headers=_headers(),
+                timeout=30,
+            )
+        except requests.RequestException:
+            time.sleep((backoff**attempt) + 0.5)
+            if attempt == max_retries - 1:
+                raise
+            continue
+
+        if response.status_code == 429:
+            retry_after = response.headers.get("Retry-After")
+            wait = (
+                float(retry_after)
+                if retry_after and retry_after.isdigit()
+                else min((backoff**attempt) + 1.0, 60.0)
+            )
+            time.sleep(wait)
+            if attempt == max_retries - 1:
+                response.raise_for_status()
+            continue
+
+        if 500 <= response.status_code < 600:
+            time.sleep(min((backoff**attempt) + 1.0, 60.0))
+            if attempt == max_retries - 1:
+                response.raise_for_status()
+            continue
+
+        response.raise_for_status()
+        return response
+    raise RuntimeError("Unreachable")
+
+
+def fetch_ticker_details(symbol: str) -> dict[str, Any]:
+    """Fetch company metadata from Polygon."""
+    response = http_get(f"{BASE}/v3/reference/tickers/{symbol}")
+    return response.json().get("results", {}) or {}
+
+
+def fetch_ohlc(symbol: str, start_date: str, end_date: str) -> list[dict[str, Any]]:
+    """Fetch daily OHLC data from Polygon."""
+    response = http_get(
+        f"{BASE}/v2/aggs/ticker/{symbol}/range/1/day/{start_date}/{end_date}",
+        params={"adjusted": "true", "sort": "asc", "limit": 50000},
+    )
+    results = response.json().get("results") or []
+    rows: list[dict[str, Any]] = []
+    for item in results:
+        rows.append(
+            {
+                "date": datetime.fromtimestamp(
+                    int(item["t"]) / 1000,
+                    tz=timezone.utc,
+                ).date().isoformat(),
+                "open": item.get("o"),
+                "high": item.get("h"),
+                "low": item.get("l"),
+                "close": item.get("c"),
+                "volume": item.get("v"),
+                "vwap": item.get("vw"),
+                "transactions": item.get("n"),
+            }
+        )
+    return rows
+
+
+def fetch_news(
+    symbol: str,
+    start_date: str,
+    end_date: str,
+    *,
+    per_page: int = 50,
+    page_sleep: float = 1.2,
+    max_pages: Optional[int] = None,
+) -> list[dict[str, Any]]:
+    """Fetch all Polygon news for a ticker, with pagination."""
+    url = f"{BASE}/v2/reference/news"
+    params = {
+        "ticker": symbol,
+        "published_utc.gte": start_date,
+        "published_utc.lte": end_date,
+        "limit": per_page,
+        "order": "asc",
+    }
+    next_url: Optional[str] = None
+    pages = 0
+    all_articles: list[dict[str, Any]] = []
+    seen_ids: set[str] = set()
+
+    while True:
+        response = http_get(next_url or url, params=None if next_url else params)
+        data = response.json()
+        results = data.get("results") or []
+        if not results:
+            break
+
+        for item in results:
+            article_id = item.get("id")
+            if article_id and article_id in seen_ids:
+                continue
+            all_articles.append(
+                {
+                    "id": article_id,
+                    "publisher": (item.get("publisher") or {}).get("name"),
+                    "title": item.get("title"),
+                    "author": item.get("author"),
+                    "published_utc": item.get("published_utc"),
+                    "amp_url": item.get("amp_url"),
+                    "article_url": item.get("article_url"),
+                    "tickers": item.get("tickers"),
+                    "description": item.get("description"),
+                    "insights": item.get("insights"),
+                }
+            )
+            if article_id:
+                seen_ids.add(article_id)
+
+        next_url = data.get("next_url")
+        pages += 1
+        if max_pages is not None and pages >= max_pages:
+            break
+        if not next_url:
+            break
+        time.sleep(page_sleep)
+
+    return all_articles
--- a/backend/data/provider_router.py
+++ b/backend/data/provider_router.py
@@ -11,7 +11,7 @@ import pandas as pd
 import yfinance as yf

 from backend.config.data_config import DataSource, get_data_sources
-from backend.data.schema import (
+from shared.schema import (
    CompanyFactsResponse,
    CompanyNews,
    CompanyNewsResponse,
@@ -30,6 +30,25 @@ logger = logging.getLogger(__name__)
 _DATA_DIR = Path(__file__).parent / "ret_data"


+def _format_provider_error(exc: Exception) -> str:
+    """Condense common provider failures into short, readable messages."""
+    message = str(exc).strip().replace("\n", " ")
+    if "429" in message:
+        return "rate limit reached"
+    if "402" in message:
+        return "insufficient credits"
+    if "422" in message or "Missing parameters" in message:
+        return "invalid request parameters"
+    if "Quote not found" in message:
+        return "quote not found"
+    return message
+
+
+def _has_valid_ticker(ticker: str) -> bool:
+    """Return whether the normalized ticker is non-empty."""
+    return bool((ticker or "").strip())
+
+
 class DataProviderRouter:
    """Route data requests across configured providers with fallbacks."""

@@ -56,6 +75,8 @@ class DataProviderRouter:
        end_date: str,
    ) -> tuple[list[Price], DataSource]:
        """Fetch prices using preferred providers with fallback."""
+        if not _has_valid_ticker(ticker):
+            return [], "local_csv"
        last_error: Optional[Exception] = None

        for source in self.price_sources():
@@ -78,7 +99,12 @@ class DataProviderRouter:
                    return prices, source
            except Exception as exc:
                last_error = exc
-                logger.warning("Price source %s failed for %s: %s", source, ticker, exc)
+                logger.warning(
+                    "Price source %s failed for %s: %s",
+                    source,
+                    ticker,
+                    _format_provider_error(exc),
+                )

        if last_error:
            raise last_error
@@ -92,6 +118,8 @@ class DataProviderRouter:
        limit: int = 10,
    ) -> tuple[list[FinancialMetrics], DataSource]:
        """Fetch financial metrics with API provider fallback."""
+        if not _has_valid_ticker(ticker):
+            return [], "local_csv"
        last_error: Optional[Exception] = None

        for source in self.api_sources():
@@ -126,7 +154,7 @@ class DataProviderRouter:
                    "Financial metrics source %s failed for %s: %s",
                    source,
                    ticker,
-                    exc,
+                    _format_provider_error(exc),
                )

        if last_error:
@@ -142,6 +170,8 @@ class DataProviderRouter:
        limit: int = 10,
    ) -> list[LineItem]:
        """Line items are only supported via Financial Datasets."""
+        if not _has_valid_ticker(ticker):
+            return []
        if "financial_datasets" not in self.api_sources():
            return []
        try:
@@ -155,7 +185,11 @@ class DataProviderRouter:
            self._record_success("line_items", "financial_datasets")
            return results
        except Exception as exc:
-            logger.warning("Line items source failed for %s: %s", ticker, exc)
+            logger.warning(
+                "Line items source failed for %s: %s",
+                ticker,
+                _format_provider_error(exc),
+            )
            return []

    def get_insider_trades(
@@ -166,6 +200,8 @@ class DataProviderRouter:
        limit: int = 1000,
    ) -> tuple[list[InsiderTrade], DataSource]:
        """Fetch insider trades with provider fallback."""
+        if not _has_valid_ticker(ticker):
+            return [], "local_csv"
        last_error: Optional[Exception] = None

        for source in self.api_sources():
@@ -193,7 +229,7 @@ class DataProviderRouter:
                    "Insider trades source %s failed for %s: %s",
                    source,
                    ticker,
-                    exc,
+                    _format_provider_error(exc),
                )

        if last_error:
@@ -208,6 +244,8 @@ class DataProviderRouter:
        limit: int = 1000,
    ) -> tuple[list[CompanyNews], DataSource]:
        """Fetch company news with provider fallback."""
+        if not _has_valid_ticker(ticker):
+            return [], "local_csv"
        last_error: Optional[Exception] = None

        for source in self.api_sources():
@@ -244,7 +282,7 @@ class DataProviderRouter:
                    "Company news source %s failed for %s: %s",
                    source,
                    ticker,
-                    exc,
+                    _format_provider_error(exc),
                )

        if last_error:
@@ -258,6 +296,8 @@ class DataProviderRouter:
        metrics_lookup,
    ) -> tuple[Optional[float], DataSource]:
        """Fetch market cap using facts API or financial metrics fallback."""
+        if not _has_valid_ticker(ticker):
+            return None, "local_csv"
        today = datetime.datetime.now().strftime("%Y-%m-%d")
        if end_date == today and "financial_datasets" in self.api_sources():
            try:
@@ -267,7 +307,7 @@ class DataProviderRouter:
                logger.warning(
                    "Market cap facts source failed for %s: %s",
                    ticker,
-                    exc,
+                    _format_provider_error(exc),
                )

        metrics, source = metrics_lookup(ticker, end_date)
--- a/backend/data/schema.py
+++ b/backend/data/schema.py
@@ -1,184 +1,50 @@
 # -*- coding: utf-8 -*-
-from pydantic import BaseModel
+"""Compatibility schema bridge.

+This module preserves the legacy ``backend.data.schema`` import path while
+delegating the actual schema definitions to ``shared.schema``. Keeping one
+canonical DTO set avoids drift as the monolith is split into service-specific
+packages.
+"""

-class Price(BaseModel):
-    open: float
-    close: float
-    high: float
-    low: float
-    volume: int
-    time: str
+from shared.schema import (
+    AgentStateData,
+    AgentStateMetadata,
+    AnalystSignal,
+    CompanyFacts,
+    CompanyFactsResponse,
+    CompanyNews,
+    CompanyNewsResponse,
+    FinancialMetrics,
+    FinancialMetricsResponse,
+    InsiderTrade,
+    InsiderTradeResponse,
+    LineItem,
+    LineItemResponse,
+    Portfolio,
+    Position,
+    Price,
+    PriceResponse,
+    TickerAnalysis,
+)

-
-class PriceResponse(BaseModel):
-    ticker: str
-    prices: list[Price]
-
-
-class FinancialMetrics(BaseModel):
-    ticker: str
-    report_period: str
-    period: str
-    currency: str
-    market_cap: float | None
-    enterprise_value: float | None
-    price_to_earnings_ratio: float | None
-    price_to_book_ratio: float | None
-    price_to_sales_ratio: float | None
-    enterprise_value_to_ebitda_ratio: float | None
-    enterprise_value_to_revenue_ratio: float | None
-    free_cash_flow_yield: float | None
-    peg_ratio: float | None
-    gross_margin: float | None
-    operating_margin: float | None
-    net_margin: float | None
-    return_on_equity: float | None
-    return_on_assets: float | None
-    return_on_invested_capital: float | None
-    asset_turnover: float | None
-    inventory_turnover: float | None
-    receivables_turnover: float | None
-    days_sales_outstanding: float | None
-    operating_cycle: float | None
-    working_capital_turnover: float | None
-    current_ratio: float | None
-    quick_ratio: float | None
-    cash_ratio: float | None
-    operating_cash_flow_ratio: float | None
-    debt_to_equity: float | None
-    debt_to_assets: float | None
-    interest_coverage: float | None
-    revenue_growth: float | None
-    earnings_growth: float | None
-    book_value_growth: float | None
-    earnings_per_share_growth: float | None
-    free_cash_flow_growth: float | None
-    operating_income_growth: float | None
-    ebitda_growth: float | None
-    payout_ratio: float | None
-    earnings_per_share: float | None
-    book_value_per_share: float | None
-    free_cash_flow_per_share: float | None
-
-
-class FinancialMetricsResponse(BaseModel):
-    financial_metrics: list[FinancialMetrics]
-
-
-class LineItem(BaseModel):
-    ticker: str
-    report_period: str
-    period: str
-    currency: str
-
-    # Allow additional fields dynamically
-    model_config = {"extra": "allow"}
-
-
-class LineItemResponse(BaseModel):
-    search_results: list[LineItem]
-
-
-class InsiderTrade(BaseModel):
-    ticker: str
-    issuer: str | None
-    name: str | None
-    title: str | None
-    is_board_director: bool | None
-    transaction_date: str | None
-    transaction_shares: float | None
-    transaction_price_per_share: float | None
-    transaction_value: float | None
-    shares_owned_before_transaction: float | None
-    shares_owned_after_transaction: float | None
-    security_title: str | None
-    filing_date: str
-
-
-class InsiderTradeResponse(BaseModel):
-    insider_trades: list[InsiderTrade]
-
-
-class CompanyNews(BaseModel):
-    category: str | None = None
-    ticker: str
-    title: str
-    related: str | None = None
-    source: str
-    date: str | None = None
-    url: str
-    summary: str | None = None
-
-
-class CompanyNewsResponse(BaseModel):
-    news: list[CompanyNews]
-
-
-class CompanyFacts(BaseModel):
-    ticker: str
-    name: str
-    cik: str | None = None
-    industry: str | None = None
-    sector: str | None = None
-    category: str | None = None
-    exchange: str | None = None
-    is_active: bool | None = None
-    listing_date: str | None = None
-    location: str | None = None
-    market_cap: float | None = None
-    number_of_employees: int | None = None
-    sec_filings_url: str | None = None
-    sic_code: str | None = None
-    sic_industry: str | None = None
-    sic_sector: str | None = None
-    website_url: str | None = None
-    weighted_average_shares: int | None = None
-
-
-class CompanyFactsResponse(BaseModel):
-    company_facts: CompanyFacts
-
-
-class Position(BaseModel):
-    """Position information - for Portfolio mode"""
-
-    long: int = 0  # Long position quantity (shares)
-    short: int = 0  # Short position quantity (shares)
-    long_cost_basis: float = 0.0  # Long position average cost
-    short_cost_basis: float = 0.0  # Short position average cost
-
-
-class Portfolio(BaseModel):
-    """Portfolio - for Portfolio mode"""
-
-    cash: float = 100000.0  # Available cash
-    positions: dict[str, Position] = {}  # ticker -> Position mapping
-    # Margin requirement (0.0 means shorting disabled, 0.5 means 50% margin)
-    margin_requirement: float = 0.0
-    margin_used: float = 0.0  # Margin used
-
-
-class AnalystSignal(BaseModel):
-    signal: str | None = None
-    confidence: float | None = None
-    reasoning: dict | str | None = None
-    max_position_size: float | None = None  # For risk management signals
-
-
-class TickerAnalysis(BaseModel):
-    ticker: str
-    analyst_signals: dict[str, AnalystSignal]  # agent_name -> signal mapping
-
-
-class AgentStateData(BaseModel):
-    tickers: list[str]
-    portfolio: Portfolio
-    start_date: str
-    end_date: str
-    ticker_analyses: dict[str, TickerAnalysis]  # ticker -> analysis mapping
-
-
-class AgentStateMetadata(BaseModel):
-    show_reasoning: bool = False
-    model_config = {"extra": "allow"}
+__all__ = [
+    "Price",
+    "PriceResponse",
+    "FinancialMetrics",
+    "FinancialMetricsResponse",
+    "LineItem",
+    "LineItemResponse",
+    "InsiderTrade",
+    "InsiderTradeResponse",
+    "CompanyNews",
+    "CompanyNewsResponse",
+    "CompanyFacts",
+    "CompanyFactsResponse",
+    "Position",
+    "Portfolio",
+    "AnalystSignal",
+    "TickerAnalysis",
+    "AgentStateData",
+    "AgentStateMetadata",
+]
--- a/backend/domains/init.py
+++ b/backend/domains/init.py
@@ -0,0 +1,2 @@
+# -*- coding: utf-8 -*-
+"""Domain modules for split service internals."""
--- a/backend/domains/news.py
+++ b/backend/domains/news.py
@@ -0,0 +1,277 @@
+# -*- coding: utf-8 -*-
+"""News/explain domain helpers shared by app surfaces and gateway fallbacks."""
+
+from __future__ import annotations
+
+from typing import Any
+
+from backend.data.market_store import MarketStore
+from backend.data.market_ingest import update_ticker_incremental
+from backend.enrich.news_enricher import enrich_news_for_symbol
+from backend.explain.range_explainer import build_range_explanation
+from backend.explain.similarity_service import find_similar_days
+from backend.explain.story_service import get_or_create_stock_story
+
+
+def news_rows_need_enrichment(rows: list[dict[str, Any]]) -> bool:
+    """Return whether news rows are missing explain-oriented analysis fields."""
+    if not rows:
+        return True
+    return all(
+        not row.get("sentiment")
+        and not row.get("relevance")
+        and not row.get("key_discussion")
+        for row in rows
+    )
+
+
+def ensure_news_fresh(
+    store: MarketStore,
+    *,
+    ticker: str,
+    target_date: str | None = None,
+) -> dict[str, Any]:
+    """Refresh raw news incrementally when stored watermarks are stale."""
+    normalized_target = str(target_date or "").strip()[:10]
+    if not normalized_target:
+        return {
+            "ticker": ticker,
+            "target_date": None,
+            "last_news_fetch": None,
+            "refreshed": False,
+        }
+
+    watermarks = store.get_ticker_watermarks(ticker)
+    last_news_fetch = str(watermarks.get("last_news_fetch") or "").strip()[:10]
+    refreshed = False
+    if not last_news_fetch or last_news_fetch < normalized_target:
+        update_ticker_incremental(
+            ticker,
+            end_date=normalized_target,
+            store=store,
+        )
+        refreshed = True
+        watermarks = store.get_ticker_watermarks(ticker)
+        last_news_fetch = str(watermarks.get("last_news_fetch") or "").strip()[:10]
+
+    return {
+        "ticker": ticker,
+        "target_date": normalized_target,
+        "last_news_fetch": last_news_fetch or None,
+        "refreshed": refreshed,
+    }
+
+
+def get_enriched_news(
+    store: MarketStore,
+    *,
+    ticker: str,
+    start_date: str | None = None,
+    end_date: str | None = None,
+    limit: int = 100,
+) -> dict[str, Any]:
+    freshness = ensure_news_fresh(store, ticker=ticker, target_date=end_date)
+    rows = store.get_news_items_enriched(
+        ticker,
+        start_date=start_date,
+        end_date=end_date,
+        limit=limit,
+    )
+    if news_rows_need_enrichment(rows):
+        enrich_news_for_symbol(
+            store,
+            ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        )
+        rows = store.get_news_items_enriched(
+            ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        )
+    return {"ticker": ticker, "news": rows, "freshness": freshness}
+
+
+def get_news_for_date(
+    store: MarketStore,
+    *,
+    ticker: str,
+    date: str,
+    limit: int = 20,
+) -> dict[str, Any]:
+    freshness = ensure_news_fresh(store, ticker=ticker, target_date=date)
+    rows = store.get_news_items_enriched(
+        ticker,
+        trade_date=date,
+        limit=limit,
+    )
+    if news_rows_need_enrichment(rows):
+        enrich_news_for_symbol(
+            store,
+            ticker,
+            start_date=date,
+            end_date=date,
+            limit=limit,
+        )
+        rows = store.get_news_items_enriched(
+            ticker,
+            trade_date=date,
+            limit=limit,
+        )
+    return {"ticker": ticker, "date": date, "news": rows, "freshness": freshness}
+
+
+def get_news_timeline(
+    store: MarketStore,
+    *,
+    ticker: str,
+    start_date: str,
+    end_date: str,
+) -> dict[str, Any]:
+    freshness = ensure_news_fresh(store, ticker=ticker, target_date=end_date)
+    timeline = store.get_news_timeline_enriched(
+        ticker,
+        start_date=start_date,
+        end_date=end_date,
+    )
+    if not timeline:
+        enrich_news_for_symbol(
+            store,
+            ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=200,
+        )
+        timeline = store.get_news_timeline_enriched(
+            ticker,
+            start_date=start_date,
+            end_date=end_date,
+        )
+    return {
+        "ticker": ticker,
+        "timeline": timeline,
+        "start_date": start_date,
+        "end_date": end_date,
+        "freshness": freshness,
+    }
+
+
+def get_news_categories(
+    store: MarketStore,
+    *,
+    ticker: str,
+    start_date: str | None = None,
+    end_date: str | None = None,
+    limit: int = 200,
+) -> dict[str, Any]:
+    freshness = ensure_news_fresh(store, ticker=ticker, target_date=end_date)
+    rows = store.get_news_items_enriched(
+        ticker,
+        start_date=start_date,
+        end_date=end_date,
+        limit=limit,
+    )
+    if news_rows_need_enrichment(rows):
+        enrich_news_for_symbol(
+            store,
+            ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        )
+    categories = store.get_news_categories_enriched(
+        ticker,
+        start_date=start_date,
+        end_date=end_date,
+        limit=limit,
+    )
+    return {"ticker": ticker, "categories": categories, "freshness": freshness}
+
+
+def get_similar_days_payload(
+    store: MarketStore,
+    *,
+    ticker: str,
+    date: str,
+    n_similar: int = 5,
+) -> dict[str, Any]:
+    freshness = ensure_news_fresh(store, ticker=ticker, target_date=date)
+    result = find_similar_days(
+        store,
+        symbol=ticker,
+        target_date=date,
+        top_k=n_similar,
+    )
+    result["freshness"] = freshness
+    return result
+
+
+def get_story_payload(
+    store: MarketStore,
+    *,
+    ticker: str,
+    as_of_date: str,
+) -> dict[str, Any]:
+    freshness = ensure_news_fresh(store, ticker=ticker, target_date=as_of_date)
+    enrich_news_for_symbol(
+        store,
+        ticker,
+        end_date=as_of_date,
+        limit=80,
+    )
+    result = get_or_create_stock_story(
+        store,
+        symbol=ticker,
+        as_of_date=as_of_date,
+    )
+    result["freshness"] = freshness
+    return result
+
+
+def get_range_explain_payload(
+    store: MarketStore,
+    *,
+    ticker: str,
+    start_date: str,
+    end_date: str,
+    article_ids: list[str] | None = None,
+    limit: int = 100,
+) -> dict[str, Any]:
+    freshness = ensure_news_fresh(store, ticker=ticker, target_date=end_date)
+    news_rows = []
+    if article_ids:
+        news_rows = store.get_news_by_ids_enriched(ticker, article_ids)
+    if not news_rows:
+        news_rows = store.get_news_items_enriched(
+            ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        )
+    if news_rows_need_enrichment(news_rows):
+        enrich_news_for_symbol(
+            store,
+            ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        )
+        news_rows = (
+            store.get_news_by_ids_enriched(ticker, article_ids)
+            if article_ids
+            else store.get_news_items_enriched(
+                ticker,
+                start_date=start_date,
+                end_date=end_date,
+                limit=limit,
+            )
+        )
+    result = build_range_explanation(
+        ticker=ticker,
+        start_date=start_date,
+        end_date=end_date,
+        news_rows=news_rows,
+    )
+    return {"ticker": ticker, "result": result, "freshness": freshness}
--- a/backend/domains/trading.py
+++ b/backend/domains/trading.py
@@ -0,0 +1,106 @@
+# -*- coding: utf-8 -*-
+"""Trading domain helpers shared by app surfaces and gateway fallbacks."""
+
+from __future__ import annotations
+
+from typing import Any
+
+from backend.services.market import MarketService
+from backend.tools.data_tools import (
+    get_company_news,
+    get_financial_metrics,
+    get_insider_trades,
+    get_market_cap,
+    get_prices,
+    search_line_items,
+)
+
+
+def get_prices_payload(*, ticker: str, start_date: str, end_date: str) -> dict[str, Any]:
+    return {
+        "ticker": ticker,
+        "prices": get_prices(ticker, start_date, end_date),
+    }
+
+
+def get_financials_payload(
+    *,
+    ticker: str,
+    end_date: str,
+    period: str = "ttm",
+    limit: int = 10,
+) -> dict[str, Any]:
+    return {
+        "financial_metrics": get_financial_metrics(
+            ticker=ticker,
+            end_date=end_date,
+            period=period,
+            limit=limit,
+        )
+    }
+
+
+def get_news_payload(
+    *,
+    ticker: str,
+    end_date: str,
+    start_date: str | None = None,
+    limit: int = 1000,
+) -> dict[str, Any]:
+    return {
+        "news": get_company_news(
+            ticker=ticker,
+            end_date=end_date,
+            start_date=start_date,
+            limit=limit,
+        )
+    }
+
+
+def get_insider_trades_payload(
+    *,
+    ticker: str,
+    end_date: str,
+    start_date: str | None = None,
+    limit: int = 1000,
+) -> dict[str, Any]:
+    return {
+        "insider_trades": get_insider_trades(
+            ticker=ticker,
+            end_date=end_date,
+            start_date=start_date,
+            limit=limit,
+        )
+    }
+
+
+def get_market_status_payload() -> dict[str, Any]:
+    market_service = MarketService(tickers=[])
+    return market_service.get_market_status()
+
+
+def get_market_cap_payload(*, ticker: str, end_date: str) -> dict[str, Any]:
+    return {
+        "ticker": ticker,
+        "end_date": end_date,
+        "market_cap": get_market_cap(ticker, end_date),
+    }
+
+
+def get_line_items_payload(
+    *,
+    ticker: str,
+    line_items: list[str],
+    end_date: str,
+    period: str = "ttm",
+    limit: int = 10,
+) -> dict[str, Any]:
+    return {
+        "search_results": search_line_items(
+            ticker=ticker,
+            line_items=line_items,
+            end_date=end_date,
+            period=period,
+            limit=limit,
+        )
+    }
--- a/backend/enrich/init.py
+++ b/backend/enrich/init.py
@@ -0,0 +1,2 @@
+"""News enrichment utilities for explain-oriented market research."""
+
--- a/backend/enrich/llm_enricher.py
+++ b/backend/enrich/llm_enricher.py
@@ -0,0 +1,301 @@
+# -*- coding: utf-8 -*-
+"""Optional AgentScope-backed news enrichment with safe local fallback."""
+
+from __future__ import annotations
+
+import asyncio
+import logging
+from concurrent.futures import ThreadPoolExecutor
+from typing import Any
+
+from pydantic import BaseModel, Field
+
+from backend.config.env_config import canonicalize_model_provider, get_env_bool, get_env_str
+from backend.llm.models import create_model
+
+logger = logging.getLogger(__name__)
+
+
+class EnrichedNewsItem(BaseModel):
+    """Structured output schema for one enriched article."""
+
+    id: str = Field(description="The source article id")
+    relevance: str = Field(description="One of high, medium, low")
+    sentiment: str = Field(description="One of positive, negative, neutral")
+    key_discussion: str = Field(description="Concise core discussion")
+    summary: str = Field(description="Concise factual summary")
+    reason_growth: str = Field(description="Growth-oriented reason if present")
+    reason_decrease: str = Field(description="Downside-oriented reason if present")
+
+
+class EnrichedNewsBatch(BaseModel):
+    """Structured output schema for a batch of enriched articles."""
+
+    items: list[EnrichedNewsItem]
+
+
+class RangeAnalysisPayload(BaseModel):
+    """Structured output schema for range explanation text."""
+
+    summary: str = Field(description="Concise Chinese range summary for the selected window")
+    trend_analysis: str = Field(description="Concise Chinese trend explanation for the selected window")
+    bullish_factors: list[str] = Field(description="Top bullish factors in Chinese")
+    bearish_factors: list[str] = Field(description="Top bearish factors in Chinese")
+
+
+def get_explain_model_info() -> dict[str, str]:
+    """Resolve provider/model used by explain enrichment."""
+    provider = canonicalize_model_provider(
+        get_env_str("EXPLAIN_ENRICH_MODEL_PROVIDER")
+        or get_env_str("MODEL_PROVIDER", "OPENAI"),
+    )
+    model_name = get_env_str("EXPLAIN_ENRICH_MODEL_NAME") or get_env_str(
+        "MODEL_NAME",
+        "gpt-4o-mini",
+    )
+    return {
+        "provider": provider,
+        "model_name": model_name,
+        "label": f"{provider}:{model_name}",
+    }
+
+
+def _normalize_enrichment_payload(payload: Any) -> dict[str, Any] | None:
+    if isinstance(payload, BaseModel):
+        payload = payload.model_dump()
+    if not isinstance(payload, dict):
+        return None
+    return {
+        "relevance": str(payload.get("relevance") or "").strip().lower() or None,
+        "sentiment": str(payload.get("sentiment") or "").strip().lower() or None,
+        "key_discussion": str(payload.get("key_discussion") or "").strip() or None,
+        "summary": str(payload.get("summary") or "").strip() or None,
+        "reason_growth": str(payload.get("reason_growth") or "").strip() or None,
+        "reason_decrease": str(payload.get("reason_decrease") or "").strip() or None,
+        "raw_json": payload,
+    }
+
+
+def _run_async(coro: Any) -> Any:
+    """Run an async AgentScope model call from sync code, even inside a running loop."""
+    try:
+        asyncio.get_running_loop()
+    except RuntimeError:
+        return asyncio.run(coro)
+
+    with ThreadPoolExecutor(max_workers=1) as executor:
+        future = executor.submit(asyncio.run, coro)
+        return future.result()
+
+
+def _get_explain_model():
+    """Create an AgentScope model for explain enrichment."""
+    model_info = get_explain_model_info()
+    return create_model(
+        model_name=model_info["model_name"],
+        provider=model_info["provider"],
+        stream=False,
+        generate_kwargs={"temperature": 0.1},
+    )
+
+
+def llm_enrichment_enabled() -> bool:
+    """Return whether AgentScope-backed LLM enrichment should be attempted."""
+    if not get_env_bool("EXPLAIN_ENRICH_USE_LLM", False):
+        return False
+    provider = get_explain_model_info()["provider"]
+    provider_key_map = {
+        "OPENAI": "OPENAI_API_KEY",
+        "ANTHROPIC": "ANTHROPIC_API_KEY",
+        "DASHSCOPE": "DASHSCOPE_API_KEY",
+        "ALIBABA": "DASHSCOPE_API_KEY",
+        "GEMINI": "GOOGLE_API_KEY",
+        "GOOGLE": "GOOGLE_API_KEY",
+        "DEEPSEEK": "DEEPSEEK_API_KEY",
+        "GROQ": "GROQ_API_KEY",
+        "OPENROUTER": "OPENROUTER_API_KEY",
+    }
+    env_key = provider_key_map.get(provider)
+    return bool(get_env_str(env_key)) if env_key else provider == "OLLAMA"
+
+
+def llm_range_analysis_enabled() -> bool:
+    """Return whether LLM range analysis should be attempted."""
+    raw_value = get_env_str("EXPLAIN_RANGE_USE_LLM")
+    if raw_value is not None and str(raw_value).strip() != "":
+        return get_env_bool("EXPLAIN_RANGE_USE_LLM", False) and llm_enrichment_enabled()
+    return llm_enrichment_enabled()
+
+
+def analyze_news_row_with_llm(row: dict[str, Any]) -> dict[str, Any] | None:
+    """Generate explain-oriented structured analysis for one article."""
+    if not llm_enrichment_enabled():
+        return None
+
+    model = _get_explain_model()
+    title = str(row.get("title") or "").strip()
+    summary = str(row.get("summary") or "").strip()
+    messages = [
+        {
+            "role": "system",
+            "content": (
+                "You produce concise structured financial news analysis. "
+                "Use only the requested fields and keep content factual."
+            ),
+        },
+        {
+            "role": "user",
+            "content": (
+                "Analyze this stock-news article for an explain UI.\n"
+                "Rules:\n"
+                "- relevance must be one of: high, medium, low\n"
+                "- sentiment must be one of: positive, negative, neutral\n"
+                "- keep each text field concise and factual\n"
+                f"- article id: {str(row.get('id') or '').strip()}\n"
+                f"Title: {title}\n"
+                f"Summary: {summary}\n"
+            ),
+        },
+    ]
+    try:
+        response = _run_async(model(messages=messages, structured_model=EnrichedNewsItem))
+    except Exception as e:
+        logger.warning(f"LLM enrichment failed: {e}")
+        return None
+
+    payload = _normalize_enrichment_payload(getattr(response, "metadata", None))
+    if payload:
+        payload.setdefault("raw_json", {})
+        payload["raw_json"]["model_provider"] = get_explain_model_info()["provider"]
+        payload["raw_json"]["model_name"] = get_explain_model_info()["model_name"]
+        payload["raw_json"]["model_label"] = get_explain_model_info()["label"]
+    return payload
+
+
+def analyze_news_rows_with_llm(rows: list[dict[str, Any]]) -> dict[str, dict[str, Any]]:
+    """Generate structured analysis for multiple articles in one request."""
+    if not llm_enrichment_enabled() or not rows:
+        return {}
+
+    payload_rows = [
+        {
+            "id": str(row.get("id") or "").strip(),
+            "title": str(row.get("title") or "").strip(),
+            "summary": str(row.get("summary") or "").strip(),
+        }
+        for row in rows
+        if str(row.get("id") or "").strip()
+    ]
+    if not payload_rows:
+        return {}
+
+    model = _get_explain_model()
+    messages = [
+        {
+            "role": "system",
+            "content": (
+                "You produce concise structured financial news analysis in JSON. "
+                "Preserve ids exactly and do not invent extra items."
+            ),
+        },
+        {
+            "role": "user",
+            "content": (
+                "Analyze these stock-news articles for an explain UI.\n"
+                "For each item return: id, relevance, sentiment, key_discussion, summary, "
+                "reason_growth, reason_decrease.\n"
+                "Rules:\n"
+                "- relevance must be one of: high, medium, low\n"
+                "- sentiment must be one of: positive, negative, neutral\n"
+                "- keep all text concise and factual\n"
+                f"Articles: {payload_rows}"
+            ),
+        },
+    ]
+    try:
+        response = _run_async(
+            model(messages=messages, structured_model=EnrichedNewsBatch),
+        )
+    except Exception:
+        return {}
+
+    metadata = getattr(response, "metadata", None)
+    if isinstance(metadata, BaseModel):
+        metadata = metadata.model_dump()
+    items = metadata.get("items") if isinstance(metadata, dict) else None
+    if not isinstance(items, list):
+        return {}
+
+    results: dict[str, dict[str, Any]] = {}
+    for item in items:
+        normalized = _normalize_enrichment_payload(item)
+        news_id = str((item.model_dump() if isinstance(item, BaseModel) else item).get("id") or "").strip() if isinstance(item, (dict, BaseModel)) else ""
+        if normalized and news_id:
+            normalized.setdefault("raw_json", {})
+            normalized["raw_json"]["model_provider"] = get_explain_model_info()["provider"]
+            normalized["raw_json"]["model_name"] = get_explain_model_info()["model_name"]
+            normalized["raw_json"]["model_label"] = get_explain_model_info()["label"]
+            results[news_id] = normalized
+    return results
+
+
+def analyze_range_with_llm(payload: dict[str, Any]) -> dict[str, Any] | None:
+    """Generate explain-oriented range summary and factor refinement."""
+    if not llm_range_analysis_enabled():
+        return None
+
+    model = _get_explain_model()
+    messages = [
+        {
+            "role": "system",
+            "content": (
+                "You write concise Chinese stock range analysis for an explain UI. "
+                "Use only the supplied facts. Keep the tone factual and analyst-like."
+            ),
+        },
+        {
+            "role": "user",
+            "content": (
+                "请基于给定事实生成区间分析。\n"
+                "输出字段：summary, trend_analysis, bullish_factors, bearish_factors。\n"
+                "要求：\n"
+                "- 全部使用简体中文\n"
+                "- summary 1到2句，概括区间走势、新闻密度和主导主题\n"
+                "- trend_analysis 1句，解释区间内部阶段变化\n"
+                "- bullish_factors 和 bearish_factors 各返回最多3条短句\n"
+                "- 不要编造未提供的信息\n"
+                f"事实数据: {payload}"
+            ),
+        },
+    ]
+    try:
+        response = _run_async(
+            model(messages=messages, structured_model=RangeAnalysisPayload),
+        )
+    except Exception as e:
+        logger.warning(f"LLM enrichment failed: {e}")
+        return None
+
+    metadata = getattr(response, "metadata", None)
+    if isinstance(metadata, BaseModel):
+        metadata = metadata.model_dump()
+    if not isinstance(metadata, dict):
+        return None
+
+    return {
+        "summary": str(metadata.get("summary") or "").strip() or None,
+        "trend_analysis": str(metadata.get("trend_analysis") or "").strip() or None,
+        "bullish_factors": [
+            str(item).strip()
+            for item in list(metadata.get("bullish_factors") or [])
+            if str(item).strip()
+        ][:3],
+        "bearish_factors": [
+            str(item).strip()
+            for item in list(metadata.get("bearish_factors") or [])
+            if str(item).strip()
+        ][:3],
+        "model_provider": get_explain_model_info()["provider"],
+        "model_name": get_explain_model_info()["model_name"],
+        "model_label": get_explain_model_info()["label"],
+    }
--- a/backend/enrich/news_enricher.py
+++ b/backend/enrich/news_enricher.py
@@ -0,0 +1,362 @@
+# -*- coding: utf-8 -*-
+"""Lightweight news enrichment for explain-oriented market analysis."""
+
+from __future__ import annotations
+
+import hashlib
+from typing import Any
+
+from backend.config.env_config import get_env_int
+from backend.enrich.llm_enricher import (
+    analyze_news_row_with_llm,
+    analyze_news_rows_with_llm,
+    llm_enrichment_enabled,
+)
+from backend.data.market_store import MarketStore
+
+
+POSITIVE_KEYWORDS = (
+    "beat", "surge", "gain", "growth", "record", "upgrade", "strong",
+    "partnership", "approved", "launch", "expands", "profit",
+)
+NEGATIVE_KEYWORDS = (
+    "miss", "drop", "fall", "cut", "downgrade", "weak", "warning",
+    "delay", "lawsuit", "probe", "tariff", "decline", "layoff",
+)
+HIGH_RELEVANCE_KEYWORDS = (
+    "earnings", "guidance", "profit", "revenue", "ceo", "fda", "tariff",
+    "regulation", "acquisition", "buyback", "forecast", "launch",
+)
+
+
+def _dedupe_key(row: dict[str, Any]) -> str:
+    trade_date = str(row.get("trade_date") or row.get("date") or "")[:10]
+    title = str(row.get("title") or "").strip().lower()
+    summary = str(row.get("summary") or "").strip().lower()[:160]
+    raw = f"{trade_date}::{title}::{summary}"
+    return hashlib.sha1(raw.encode("utf-8")).hexdigest()
+
+
+def _chunk_rows(rows: list[dict[str, Any]], size: int) -> list[list[dict[str, Any]]]:
+    chunk_size = max(1, int(size))
+    return [rows[index:index + chunk_size] for index in range(0, len(rows), chunk_size)]
+
+
+def classify_news_row(row: dict[str, Any]) -> dict[str, Any]:
+    """Return a lightweight explain-oriented analysis for one article."""
+    llm_result = analyze_news_row_with_llm(row)
+    if isinstance(llm_result, dict):
+        merged = dict(llm_result)
+        merged.setdefault("summary", str(row.get("summary") or row.get("title") or "")[:280])
+        merged.setdefault("raw_json", row)
+        merged["analysis_source"] = "llm"
+        return merged
+
+    title = str(row.get("title") or "").strip()
+    summary = str(row.get("summary") or "").strip()
+    text = f"{title} {summary}".lower()
+
+    positive_hits = [keyword for keyword in POSITIVE_KEYWORDS if keyword in text]
+    negative_hits = [keyword for keyword in NEGATIVE_KEYWORDS if keyword in text]
+    relevance_hits = [keyword for keyword in HIGH_RELEVANCE_KEYWORDS if keyword in text]
+
+    if len(positive_hits) > len(negative_hits):
+        sentiment = "positive"
+    elif len(negative_hits) > len(positive_hits):
+        sentiment = "negative"
+    else:
+        sentiment = "neutral"
+
+    relevance = "high" if relevance_hits else "medium" if title else "low"
+    summary_text = summary or title
+    key_discussion = ""
+    if relevance_hits:
+        key_discussion = f"核心主题集中在 {', '.join(relevance_hits[:3])}"
+    elif summary_text:
+        key_discussion = summary_text[:160]
+
+    reason_growth = ""
+    reason_decrease = ""
+    if sentiment == "positive":
+        reason_growth = summary_text[:200]
+    elif sentiment == "negative":
+        reason_decrease = summary_text[:200]
+
+    return {
+        "relevance": relevance,
+        "sentiment": sentiment,
+        "key_discussion": key_discussion,
+        "summary": summary_text[:280],
+        "reason_growth": reason_growth,
+        "reason_decrease": reason_decrease,
+        "analysis_source": "local",
+        "raw_json": row,
+    }
+
+
+def attach_forward_returns(
+    *,
+    news_rows: list[dict[str, Any]],
+    ohlc_rows: list[dict[str, Any]],
+) -> list[dict[str, Any]]:
+    """Attach forward-return labels to each analyzed row."""
+    if not ohlc_rows:
+        return news_rows
+
+    closes_by_date = {
+        str(row.get("date")): float(row.get("close"))
+        for row in ohlc_rows
+        if row.get("date") is not None and row.get("close") is not None
+    }
+    ordered_dates = [str(row.get("date")) for row in ohlc_rows if row.get("date") is not None]
+    date_index = {date: idx for idx, date in enumerate(ordered_dates)}
+
+    horizons = {
+        "ret_t0": 0,
+        "ret_t1": 1,
+        "ret_t3": 3,
+        "ret_t5": 5,
+        "ret_t10": 10,
+    }
+
+    enriched: list[dict[str, Any]] = []
+    for row in news_rows:
+        trade_date = str(row.get("trade_date") or "")[:10]
+        base_close = closes_by_date.get(trade_date)
+        if not trade_date or base_close in (None, 0):
+            enriched.append(row)
+            continue
+
+        next_row = dict(row)
+        base_index = date_index.get(trade_date)
+        if base_index is None:
+            enriched.append(next_row)
+            continue
+
+        for field, offset in horizons.items():
+            target_index = base_index + offset
+            if target_index >= len(ordered_dates):
+                next_row[field] = None
+                continue
+            target_close = closes_by_date.get(ordered_dates[target_index])
+            next_row[field] = (
+                (float(target_close) - float(base_close)) / float(base_close)
+                if target_close not in (None, 0)
+                else None
+            )
+        enriched.append(next_row)
+    return enriched
+
+
+def build_analysis_rows(
+    *,
+    symbol: str,
+    news_rows: list[dict[str, Any]],
+    ohlc_rows: list[dict[str, Any]],
+) -> tuple[list[dict[str, Any]], dict[str, int]]:
+    """Transform raw news rows into market_store news_analysis payloads plus stats."""
+    llm_results: dict[str, dict[str, Any]] = {}
+    if llm_enrichment_enabled():
+        batch_size = get_env_int("EXPLAIN_ENRICH_BATCH_SIZE", 8)
+        for chunk in _chunk_rows(news_rows, batch_size):
+            llm_results.update(analyze_news_rows_with_llm(chunk))
+
+    staged_rows: list[dict[str, Any]] = []
+    seen_dedupe_keys: set[str] = set()
+    deduped_count = 0
+    llm_count = 0
+    local_count = 0
+    for row in news_rows:
+        news_id = str(row.get("id") or "").strip()
+        if not news_id:
+            continue
+        dedupe_key = _dedupe_key(row)
+        if dedupe_key in seen_dedupe_keys:
+            deduped_count += 1
+            continue
+        seen_dedupe_keys.add(dedupe_key)
+        batch_result = llm_results.get(news_id)
+        if isinstance(batch_result, dict):
+            analysis = dict(batch_result)
+            analysis.setdefault("summary", str(row.get("summary") or row.get("title") or "")[:280])
+            analysis.setdefault("raw_json", row)
+            analysis["analysis_source"] = "llm"
+            llm_count += 1
+        else:
+            analysis = classify_news_row(row)
+            if analysis.get("analysis_source") == "llm":
+                llm_count += 1
+            else:
+                local_count += 1
+        staged_rows.append(
+            {
+                "news_id": news_id,
+                "trade_date": str(row.get("trade_date") or "")[:10] or None,
+                **analysis,
+            }
+        )
+    return (
+        attach_forward_returns(news_rows=staged_rows, ohlc_rows=ohlc_rows),
+        {
+            "deduped_count": deduped_count,
+            "llm_count": llm_count,
+            "local_count": local_count,
+        },
+    )
+
+
+def enrich_news_for_symbol(
+    store: MarketStore,
+    symbol: str,
+    *,
+    start_date: str | None = None,
+    end_date: str | None = None,
+    limit: int = 200,
+    analysis_source: str = "local",
+    skip_existing: bool = True,
+    only_reanalyze_local: bool = False,
+) -> dict[str, Any]:
+    """Read raw market news, compute explain fields, and persist them."""
+    normalized_symbol = str(symbol or "").strip().upper()
+    if not normalized_symbol:
+        return {"symbol": "", "analyzed": 0}
+
+    news_rows = store.get_news_items(
+        normalized_symbol,
+        start_date=start_date,
+        end_date=end_date,
+        limit=limit,
+    )
+    total_news_count = len(news_rows)
+    skipped_existing_count = 0
+    analyzed_sources: dict[str, str] = {}
+    skipped_missing_analysis_count = 0
+    skipped_non_local_count = 0
+    if news_rows and only_reanalyze_local:
+        analyzed_sources = store.get_analyzed_news_sources(
+            normalized_symbol,
+            start_date=start_date,
+            end_date=end_date,
+        )
+        skipped_missing_analysis_count = sum(
+            1
+            for row in news_rows
+            if str(row.get("id") or "").strip() not in analyzed_sources
+        )
+        skipped_non_local_count = sum(
+            1
+            for row in news_rows
+            if str(row.get("id") or "").strip() in analyzed_sources
+            and analyzed_sources.get(str(row.get("id") or "").strip()) != "local"
+        )
+        skipped_existing_count = sum(
+            1
+            for row in news_rows
+            if str(row.get("id") or "").strip() not in analyzed_sources
+            or analyzed_sources.get(str(row.get("id") or "").strip()) != "local"
+        )
+        news_rows = [
+            row for row in news_rows
+            if analyzed_sources.get(str(row.get("id") or "").strip()) == "local"
+        ]
+    elif skip_existing and news_rows:
+        analyzed_ids = store.get_analyzed_news_ids(
+            normalized_symbol,
+            start_date=start_date,
+            end_date=end_date,
+        )
+        skipped_existing_count = sum(
+            1
+            for row in news_rows
+            if str(row.get("id") or "").strip() in analyzed_ids
+        )
+        news_rows = [
+            row for row in news_rows
+            if str(row.get("id") or "").strip() not in analyzed_ids
+        ]
+    ohlc_start = start_date or (news_rows[-1]["trade_date"] if news_rows and news_rows[-1].get("trade_date") else None)
+    ohlc_end = end_date or (news_rows[0]["trade_date"] if news_rows and news_rows[0].get("trade_date") else None)
+    ohlc_rows = (
+        store.get_ohlc(normalized_symbol, ohlc_start, ohlc_end)
+        if ohlc_start and ohlc_end
+        else []
+    )
+    analysis_rows, stats = build_analysis_rows(
+        symbol=normalized_symbol,
+        news_rows=news_rows,
+        ohlc_rows=ohlc_rows,
+    )
+    analyzed = store.upsert_news_analysis(
+        normalized_symbol,
+        analysis_rows,
+        analysis_source=analysis_source,
+    )
+    upgraded_dates = sorted(
+        {
+            str(row.get("trade_date") or "")[:10]
+            for row in analysis_rows
+            if str(row.get("analysis_source") or "").strip().lower() == "llm"
+            and str(row.get("trade_date") or "").strip()
+        }
+    )
+    remaining_local_titles = [
+        str(row.get("title") or row.get("news_id") or "").strip()
+        for row in news_rows
+        for analyzed_row in analysis_rows
+        if str(analyzed_row.get("news_id") or "").strip() == str(row.get("id") or "").strip()
+        and str(analyzed_row.get("analysis_source") or "").strip().lower() == "local"
+    ][:5]
+    return {
+        "symbol": normalized_symbol,
+        "analyzed": analyzed,
+        "news_count": total_news_count,
+        "queued_count": len(news_rows),
+        "skipped_existing_count": skipped_existing_count,
+        "deduped_count": stats["deduped_count"],
+        "llm_count": stats["llm_count"],
+        "local_count": stats["local_count"],
+        "only_reanalyze_local": only_reanalyze_local,
+        "upgraded_local_to_llm_count": (
+            stats["llm_count"]
+            if only_reanalyze_local
+            else 0
+        ),
+        "execution_summary": {
+            "upgraded_dates": upgraded_dates[:5],
+            "remaining_local_titles": remaining_local_titles,
+            "skipped_missing_analysis_count": skipped_missing_analysis_count,
+            "skipped_non_local_count": skipped_non_local_count,
+        },
+    }
+
+
+def enrich_symbols(
+    store: MarketStore,
+    symbols: list[str],
+    *,
+    start_date: str | None = None,
+    end_date: str | None = None,
+    limit: int = 200,
+    analysis_source: str = "local",
+    skip_existing: bool = True,
+    only_reanalyze_local: bool = False,
+) -> list[dict[str, Any]]:
+    """Batch enrich multiple symbols for explain-oriented news analysis."""
+    results = []
+    for symbol in symbols:
+        normalized_symbol = str(symbol or "").strip().upper()
+        if not normalized_symbol:
+            continue
+        results.append(
+            enrich_news_for_symbol(
+                store,
+                normalized_symbol,
+                start_date=start_date,
+                end_date=end_date,
+                limit=limit,
+                analysis_source=analysis_source,
+                skip_existing=skip_existing,
+                only_reanalyze_local=only_reanalyze_local,
+            )
+        )
+    return results
--- a/backend/explain/init.py
+++ b/backend/explain/init.py
@@ -0,0 +1,2 @@
+# -*- coding: utf-8 -*-
+"""Explain-oriented services for stock narratives and news research."""
--- a/backend/explain/category_engine.py
+++ b/backend/explain/category_engine.py
@@ -0,0 +1,69 @@
+# -*- coding: utf-8 -*-
+"""Rule-based news categorization for explain UI."""
+
+from __future__ import annotations
+
+from typing import Any, Dict, Iterable
+
+
+CATEGORY_KEYWORDS = {
+    "market": [
+        "market", "stock", "rally", "sell-off", "selloff", "trading",
+        "wall street", "s&p", "nasdaq", "dow", "index", "bull", "bear",
+        "correction", "volatility",
+    ],
+    "policy": [
+        "regulation", "fed", "federal reserve", "tariff", "sanction",
+        "interest rate", "policy", "government", "congress", "sec",
+        "trade war", "ban", "legislation", "tax",
+    ],
+    "earnings": [
+        "earnings", "revenue", "profit", "quarter", "eps", "guidance",
+        "forecast", "income", "sales", "beat", "miss", "outlook",
+        "financial results",
+    ],
+    "product_tech": [
+        "product", "ai", "chip", "cloud", "launch", "patent",
+        "technology", "innovation", "release", "platform", "model",
+        "software", "hardware", "gpu", "autonomous",
+    ],
+    "competition": [
+        "competitor", "rival", "market share", "overtake", "compete",
+        "competition", "vs", "versus", "battle", "challenge",
+    ],
+    "management": [
+        "ceo", "executive", "resign", "layoff", "restructure",
+        "management", "leadership", "appoint", "hire", "board",
+        "chairman",
+    ],
+}
+
+
+def categorize_news_rows(rows: Iterable[Dict[str, Any]]) -> Dict[str, Dict[str, Any]]:
+    """Bucket news rows by keyword categories."""
+    categories: Dict[str, Dict[str, Any]] = {
+        key: {
+            "label": key,
+            "count": 0,
+            "article_ids": [],
+        }
+        for key in CATEGORY_KEYWORDS
+    }
+
+    for row in rows:
+        text = " ".join(
+            [
+                str(row.get("title") or ""),
+                str(row.get("summary") or ""),
+                str(row.get("related") or ""),
+                str(row.get("category") or ""),
+            ]
+        ).lower()
+        article_id = row.get("id")
+        for category, keywords in CATEGORY_KEYWORDS.items():
+            if any(keyword in text for keyword in keywords):
+                categories[category]["count"] += 1
+                if article_id:
+                    categories[category]["article_ids"].append(article_id)
+
+    return categories
--- a/backend/explain/range_explainer.py
+++ b/backend/explain/range_explainer.py
@@ -0,0 +1,214 @@
+# -*- coding: utf-8 -*-
+"""Local range explanation built from price and persisted news."""
+
+from __future__ import annotations
+
+from typing import Any, Dict
+
+from backend.enrich.llm_enricher import analyze_range_with_llm
+from backend.explain.category_engine import categorize_news_rows
+from backend.tools.data_tools import get_prices
+
+
+def _rank_event_score(row: Dict[str, Any]) -> float:
+    relevance = str(row.get("relevance") or "").strip().lower()
+    relevance_score = {"high": 3.0, "relevant": 3.0, "medium": 2.0, "low": 1.0}.get(
+        relevance,
+        0.5,
+    )
+    impact_score = abs(float(row.get("ret_t0") or 0.0)) * 100
+    return relevance_score + impact_score
+
+
+def summarize_bullish_factors(
+    news_rows: list[Dict[str, Any]],
+    *,
+    limit: int = 5,
+) -> list[str]:
+    factors = []
+    for row in news_rows:
+        if str(row.get("sentiment") or "").strip().lower() != "positive":
+            continue
+        candidate = row.get("reason_growth") or row.get("key_discussion") or row.get("summary") or row.get("title")
+        if candidate:
+            factors.append(str(candidate).strip())
+    seen = set()
+    output = []
+    for factor in factors:
+        if factor in seen:
+            continue
+        seen.add(factor)
+        output.append(factor[:200])
+        if len(output) >= limit:
+            break
+    return output
+
+
+def summarize_bearish_factors(
+    news_rows: list[Dict[str, Any]],
+    *,
+    limit: int = 5,
+) -> list[str]:
+    factors = []
+    for row in news_rows:
+        if str(row.get("sentiment") or "").strip().lower() != "negative":
+            continue
+        candidate = row.get("reason_decrease") or row.get("key_discussion") or row.get("summary") or row.get("title")
+        if candidate:
+            factors.append(str(candidate).strip())
+    seen = set()
+    output = []
+    for factor in factors:
+        if factor in seen:
+            continue
+        seen.add(factor)
+        output.append(factor[:200])
+        if len(output) >= limit:
+            break
+    return output
+
+
+def build_trend_analysis(prices: list[Any]) -> str:
+    if len(prices) < 2:
+        return "区间样本较短，暂不具备足够趋势信息。"
+    if len(prices) < 3:
+        open_price = float(prices[0].open)
+        close_price = float(prices[-1].close)
+        change = ((close_price - open_price) / open_price) * 100 if open_price else 0.0
+        return f"短区间内价格变动 {change:+.2f}%，趋势信息有限。"
+
+    mid = len(prices) // 2
+    first_open = float(prices[0].open)
+    first_close = float(prices[mid].close)
+    second_open = float(prices[mid].open)
+    second_close = float(prices[-1].close)
+    first_half = ((first_close - first_open) / first_open) * 100 if first_open else 0.0
+    second_half = ((second_close - second_open) / second_open) * 100 if second_open else 0.0
+    return (
+        f"前半段{'上涨' if first_half >= 0 else '下跌'} {abs(first_half):.2f}%，"
+        f"后半段{'上涨' if second_half >= 0 else '下跌'} {abs(second_half):.2f}%，"
+        "说明价格驱动在区间内部出现了阶段性切换。"
+    )
+
+
+def build_range_explanation(
+    *,
+    ticker: str,
+    start_date: str,
+    end_date: str,
+    news_rows: list[Dict[str, Any]],
+) -> Dict[str, Any]:
+    """Explain a price range with local price and news heuristics."""
+    prices = get_prices(ticker, start_date, end_date)
+    if not prices:
+        return {
+            "symbol": ticker,
+            "start_date": start_date,
+            "end_date": end_date,
+            "error": "No OHLC data for this range",
+        }
+
+    open_price = float(prices[0].open)
+    close_price = float(prices[-1].close)
+    high_price = max(float(price.high) for price in prices)
+    low_price = min(float(price.low) for price in prices)
+    total_volume = sum(int(price.volume) for price in prices)
+    price_change_pct = (
+        ((close_price - open_price) / open_price) * 100 if open_price else 0.0
+    )
+
+    categories = categorize_news_rows(news_rows)
+    news_count = len(news_rows)
+    dominant_categories = sorted(
+        (
+            {"category": key, "count": value["count"]}
+            for key, value in categories.items()
+            if value["count"] > 0
+        ),
+        key=lambda item: item["count"],
+        reverse=True,
+    )
+
+    direction = "上涨" if price_change_pct > 0 else "下跌" if price_change_pct < 0 else "横盘"
+    category_text = (
+        f"主要主题集中在 {', '.join(item['category'] for item in dominant_categories[:3])}。"
+        if dominant_categories
+        else "区间内未识别出明显的主题聚类。"
+    )
+    summary = (
+        f"{ticker} 在 {start_date} 至 {end_date} 区间内{direction} {abs(price_change_pct):.2f}%，"
+        f"区间覆盖 {len(prices)} 个交易日，关联新闻 {news_count} 条。{category_text}"
+    )
+
+    bullish_factors = summarize_bullish_factors(news_rows)
+    bearish_factors = summarize_bearish_factors(news_rows)
+    trend_analysis = build_trend_analysis(prices)
+    llm_source = "local"
+
+    range_payload = {
+        "ticker": ticker,
+        "start_date": start_date,
+        "end_date": end_date,
+        "price_change_pct": round(price_change_pct, 2),
+        "trading_days": len(prices),
+        "news_count": news_count,
+        "dominant_categories": dominant_categories[:5],
+        "bullish_factors": bullish_factors[:3],
+        "bearish_factors": bearish_factors[:3],
+        "trend_analysis": trend_analysis,
+        "top_news": [
+            {
+                "date": row.get("trade_date") or str(row.get("date") or "")[:10],
+                "title": row.get("title") or "",
+                "summary": row.get("summary") or "",
+                "sentiment": row.get("sentiment") or "",
+                "relevance": row.get("relevance") or "",
+                "ret_t0": row.get("ret_t0"),
+            }
+            for row in sorted(news_rows, key=_rank_event_score, reverse=True)[:5]
+        ],
+    }
+    llm_analysis = analyze_range_with_llm(range_payload)
+    if isinstance(llm_analysis, dict):
+        summary = llm_analysis.get("summary") or summary
+        trend_analysis = llm_analysis.get("trend_analysis") or trend_analysis
+        bullish_factors = llm_analysis.get("bullish_factors") or bullish_factors
+        bearish_factors = llm_analysis.get("bearish_factors") or bearish_factors
+        llm_source = "llm"
+
+    key_events = [
+        {
+            "date": row.get("trade_date") or str(row.get("date") or "")[:10],
+            "title": row.get("title") or "Untitled news",
+            "summary": row.get("summary") or "",
+            "category": row.get("category") or "",
+            "id": row.get("id"),
+            "sentiment": row.get("sentiment"),
+            "ret_t0": row.get("ret_t0"),
+        }
+        for row in sorted(news_rows, key=_rank_event_score, reverse=True)[:8]
+    ]
+
+    return {
+        "symbol": ticker,
+        "start_date": start_date,
+        "end_date": end_date,
+        "price_change_pct": round(price_change_pct, 2),
+        "open_price": open_price,
+        "close_price": close_price,
+        "high_price": high_price,
+        "low_price": low_price,
+        "total_volume": total_volume,
+        "trading_days": len(prices),
+        "news_count": news_count,
+        "dominant_categories": dominant_categories[:5],
+        "analysis": {
+            "summary": summary,
+            "key_events": key_events,
+            "bullish_factors": bullish_factors,
+            "bearish_factors": bearish_factors,
+            "trend_analysis": trend_analysis,
+            "analysis_source": llm_source,
+            "analysis_model_label": llm_analysis.get("model_label") if isinstance(llm_analysis, dict) else None,
+        },
+    }
--- a/backend/explain/similarity_service.py
+++ b/backend/explain/similarity_service.py
@@ -0,0 +1,202 @@
+# -*- coding: utf-8 -*-
+"""Same-ticker historical similar day search for explain view."""
+
+from __future__ import annotations
+
+from math import sqrt
+from typing import Any
+
+from backend.data.market_store import MarketStore
+
+
+def _safe_float(value: Any, default: float = 0.0) -> float:
+    try:
+        parsed = float(value)
+    except (TypeError, ValueError):
+        return default
+    return parsed
+
+
+def build_daily_feature_rows(
+    *,
+    symbol: str,
+    ohlc_rows: list[dict[str, Any]],
+    news_rows: list[dict[str, Any]],
+) -> list[dict[str, Any]]:
+    """Aggregate price/news context into daily feature rows."""
+    price_by_date = {str(row.get("date")): row for row in ohlc_rows if row.get("date")}
+    ordered_dates = [str(row.get("date")) for row in ohlc_rows if row.get("date")]
+
+    news_by_date: dict[str, list[dict[str, Any]]] = {}
+    for row in news_rows:
+        trade_date = str(row.get("trade_date") or "")[:10] or str(row.get("date") or "")[:10]
+        if not trade_date:
+            continue
+        news_by_date.setdefault(trade_date, []).append(row)
+
+    features: list[dict[str, Any]] = []
+    previous_close: float | None = None
+    for idx, date in enumerate(ordered_dates):
+        price_row = price_by_date[date]
+        close_price = _safe_float(price_row.get("close"))
+        open_price = _safe_float(price_row.get("open"), close_price)
+        day_news = news_by_date.get(date, [])
+        positive_count = sum(1 for item in day_news if str(item.get("sentiment") or "").lower() == "positive")
+        negative_count = sum(1 for item in day_news if str(item.get("sentiment") or "").lower() == "negative")
+        high_relevance_count = sum(
+            1 for item in day_news if str(item.get("relevance") or "").lower() in {"high", "relevant"}
+        )
+        ret_1d = (
+            ((close_price - previous_close) / previous_close)
+            if previous_close not in (None, 0)
+            else 0.0
+        )
+        intraday_ret = ((close_price - open_price) / open_price) if open_price else 0.0
+        sentiment_score = (
+            (positive_count - negative_count) / max(len(day_news), 1)
+            if day_news
+            else 0.0
+        )
+        future_t1 = None
+        future_t3 = None
+        if idx + 1 < len(ordered_dates) and close_price:
+            next_close = _safe_float(price_by_date[ordered_dates[idx + 1]].get("close"))
+            future_t1 = ((next_close - close_price) / close_price) if next_close else None
+        if idx + 3 < len(ordered_dates) and close_price:
+            next_close = _safe_float(price_by_date[ordered_dates[idx + 3]].get("close"))
+            future_t3 = ((next_close - close_price) / close_price) if next_close else None
+
+        features.append(
+            {
+                "date": date,
+                "symbol": symbol,
+                "n_articles": len(day_news),
+                "positive_count": positive_count,
+                "negative_count": negative_count,
+                "high_relevance_count": high_relevance_count,
+                "sentiment_score": sentiment_score,
+                "ret_1d": ret_1d,
+                "intraday_ret": intraday_ret,
+                "close": close_price,
+                "ret_t1_after": future_t1,
+                "ret_t3_after": future_t3,
+                "news": [
+                    {
+                        "title": row.get("title") or "",
+                        "sentiment": row.get("sentiment") or "neutral",
+                    }
+                    for row in day_news[:3]
+                ],
+            }
+        )
+        previous_close = close_price
+    return features
+
+
+def compute_similarity_scores(
+    target_vector: list[float],
+    candidate_vectors: list[tuple[str, list[float], dict[str, Any]]],
+) -> list[dict[str, Any]]:
+    """Return sorted similarity matches based on normalized Euclidean distance."""
+    if not candidate_vectors:
+        return []
+    dimensions = len(target_vector)
+    ranges = []
+    for dimension in range(dimensions):
+        values = [vector[1][dimension] for vector in candidate_vectors] + [target_vector[dimension]]
+        min_value = min(values)
+        max_value = max(values)
+        ranges.append(max(max_value - min_value, 1e-9))
+
+    scored = []
+    for date, vector, payload in candidate_vectors:
+        distance = sqrt(
+            sum(
+                ((target_vector[i] - vector[i]) / ranges[i]) ** 2
+                for i in range(dimensions)
+            )
+        )
+        similarity = 1.0 / (1.0 + distance)
+        scored.append(
+            {
+                "date": date,
+                "score": round(similarity, 4),
+                **payload,
+            }
+        )
+    return sorted(scored, key=lambda item: item["score"], reverse=True)
+
+
+def find_similar_days(
+    store: MarketStore,
+    *,
+    symbol: str,
+    target_date: str,
+    top_k: int = 10,
+) -> dict[str, Any]:
+    """Find same-ticker historical days most similar to a target day."""
+    cached = store.get_similar_day_cache(symbol, target_date=target_date)
+    if cached and cached.get("payload"):
+        return cached["payload"]
+
+    ohlc_rows = store.get_ohlc(symbol, "1900-01-01", target_date)
+    news_rows = store.get_news_items_enriched(symbol, end_date=target_date, limit=500)
+    daily_rows = build_daily_feature_rows(symbol=symbol, ohlc_rows=ohlc_rows, news_rows=news_rows)
+    feature_map = {row["date"]: row for row in daily_rows}
+    target_row = feature_map.get(target_date)
+    if not target_row:
+        return {
+            "symbol": symbol,
+            "target_date": target_date,
+            "items": [],
+            "error": "No feature row for target date",
+        }
+
+    vector_keys = [
+        "sentiment_score",
+        "n_articles",
+        "positive_count",
+        "negative_count",
+        "high_relevance_count",
+        "ret_1d",
+        "intraday_ret",
+    ]
+    target_vector = [_safe_float(target_row.get(key)) for key in vector_keys]
+    candidates = []
+    for row in daily_rows:
+        date = row["date"]
+        if date == target_date:
+            continue
+        payload = {
+            "n_articles": row["n_articles"],
+            "sentiment_score": round(row["sentiment_score"], 4),
+            "ret_1d": round(row["ret_1d"] * 100, 2),
+            "intraday_ret": round(row["intraday_ret"] * 100, 2),
+            "ret_t1_after": round(row["ret_t1_after"] * 100, 2) if row["ret_t1_after"] is not None else None,
+            "ret_t3_after": round(row["ret_t3_after"] * 100, 2) if row["ret_t3_after"] is not None else None,
+            "top_reasons": [item["title"] for item in row["news"][:2] if item.get("title")],
+            "news": row["news"],
+        }
+        candidates.append(
+            (
+                date,
+                [_safe_float(row.get(key)) for key in vector_keys],
+                payload,
+            )
+        )
+
+    items = compute_similarity_scores(target_vector, candidates)[: max(1, min(int(top_k), 20))]
+    result = {
+        "symbol": symbol,
+        "target_date": target_date,
+        "target_features": {
+            "sentiment_score": round(target_row["sentiment_score"], 4),
+            "n_articles": target_row["n_articles"],
+            "ret_1d": round(target_row["ret_1d"] * 100, 2),
+            "intraday_ret": round(target_row["intraday_ret"] * 100, 2),
+            "high_relevance_count": target_row["high_relevance_count"],
+        },
+        "items": items,
+    }
+    store.upsert_similar_day_cache(symbol, target_date=target_date, payload=result, source="local")
+    return result
--- a/backend/explain/story_service.py
+++ b/backend/explain/story_service.py
@@ -0,0 +1,127 @@
+# -*- coding: utf-8 -*-
+"""Stock story generation for explain view."""
+
+from __future__ import annotations
+
+from datetime import datetime, timedelta
+from typing import Any
+
+from backend.data.market_store import MarketStore
+
+
+def build_stock_story(
+    *,
+    symbol: str,
+    as_of_date: str,
+    price_rows: list[dict[str, Any]],
+    news_rows: list[dict[str, Any]],
+) -> str:
+    """Build a compact markdown story from enriched news and recent price action."""
+    lines = [f"## {symbol} Story", f"As of `{as_of_date}`"]
+    if not price_rows:
+        lines.append("")
+        lines.append("No OHLC data available for story generation.")
+        return "\n".join(lines)
+
+    open_price = float(price_rows[0].get("open") or price_rows[0].get("close") or 0.0)
+    close_price = float(price_rows[-1].get("close") or 0.0)
+    price_change = ((close_price - open_price) / open_price) * 100 if open_price else 0.0
+    high_price = max(float(row.get("high") or row.get("close") or 0.0) for row in price_rows)
+    low_price = min(float(row.get("low") or row.get("close") or 0.0) for row in price_rows)
+
+    lines.append("")
+    lines.append(
+        f"The stock moved {'up' if price_change >= 0 else 'down'} "
+        f"{abs(price_change):.2f}% over the recent window, trading between "
+        f"${low_price:.2f} and ${high_price:.2f}."
+    )
+
+    positive = [row for row in news_rows if str(row.get("sentiment") or "").lower() == "positive"]
+    negative = [row for row in news_rows if str(row.get("sentiment") or "").lower() == "negative"]
+    lines.append("")
+    lines.append(
+        f"Recent coverage included {len(news_rows)} relevant articles "
+        f"({len(positive)} positive / {len(negative)} negative)."
+    )
+
+    if news_rows:
+        lines.append("")
+        lines.append("### Key Moments")
+        ranked_rows = sorted(
+            news_rows,
+            key=lambda row: (
+                0 if str(row.get("relevance") or "").lower() in {"high", "relevant"} else 1,
+                -abs(float(row.get("ret_t0") or 0.0)),
+            ),
+        )
+        for row in ranked_rows[:5]:
+            trade_date = row.get("trade_date") or str(row.get("date") or "")[:10]
+            title = row.get("title") or "Untitled"
+            key_discussion = row.get("key_discussion") or row.get("summary") or ""
+            sentiment = str(row.get("sentiment") or "neutral").lower()
+            lines.append(
+                f"- `{trade_date}` [{sentiment}] {title}: {str(key_discussion).strip()[:220]}"
+            )
+
+    if positive:
+        lines.append("")
+        lines.append("### Bullish Threads")
+        for row in positive[:3]:
+            reason = row.get("reason_growth") or row.get("key_discussion") or row.get("summary") or row.get("title")
+            lines.append(f"- {str(reason).strip()[:220]}")
+
+    if negative:
+        lines.append("")
+        lines.append("### Bearish Threads")
+        for row in negative[:3]:
+            reason = row.get("reason_decrease") or row.get("key_discussion") or row.get("summary") or row.get("title")
+            lines.append(f"- {str(reason).strip()[:220]}")
+
+    return "\n".join(lines)
+
+
+def get_or_create_stock_story(
+    store: MarketStore,
+    *,
+    symbol: str,
+    as_of_date: str,
+) -> dict[str, Any]:
+    """Return cached story or build a new one from recent market context."""
+    cached = store.get_story_cache(symbol, as_of_date=as_of_date)
+    if cached:
+        return {
+            "symbol": symbol,
+            "as_of_date": as_of_date,
+            "story": cached.get("content") or "",
+            "source": cached.get("source") or "cache",
+        }
+
+    start_date = None
+    if len(as_of_date) >= 10:
+        target_date = datetime.strptime(as_of_date[:10], "%Y-%m-%d").date()
+        start_date = (target_date - timedelta(days=29)).isoformat()
+
+    price_rows = (
+        store.get_ohlc(symbol, start_date, as_of_date)
+        if start_date
+        else []
+    )
+    news_rows = store.get_news_items_enriched(
+        symbol,
+        start_date=start_date,
+        end_date=as_of_date,
+        limit=40,
+    )
+    story = build_stock_story(
+        symbol=symbol,
+        as_of_date=as_of_date,
+        price_rows=price_rows,
+        news_rows=news_rows,
+    )
+    store.upsert_story_cache(symbol, as_of_date=as_of_date, content=story, source="local")
+    return {
+        "symbol": symbol,
+        "as_of_date": as_of_date,
+        "story": story,
+        "source": "local",
+    }
--- a/backend/gateway_server.py
+++ b/backend/gateway_server.py
@@ -0,0 +1,251 @@
+# -*- coding: utf-8 -*-
+"""Gateway Server - Entry point for Gateway subprocess.
+
+This module is launched as a subprocess by the Control Plane (FastAPI)
+to run the Data Plane (Gateway + Pipeline).
+"""
+
+import argparse
+import asyncio
+import json
+import logging
+import os
+import sys
+from contextlib import AsyncExitStack
+from pathlib import Path
+
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+from backend.agents import AnalystAgent, PMAgent, RiskAgent
+from backend.agents.skills_manager import SkillsManager
+from backend.agents.toolkit_factory import create_agent_toolkit, load_agent_profiles
+from backend.agents.prompt_loader import get_prompt_loader
+from backend.agents.workspace_manager import WorkspaceManager
+from backend.config.constants import ANALYST_TYPES
+from backend.core.pipeline import TradingPipeline
+from backend.core.pipeline_runner import create_agents, create_long_term_memory
+from backend.core.scheduler import BacktestScheduler, Scheduler
+from backend.llm.models import get_agent_formatter, get_agent_model
+from backend.runtime.manager import (
+    TradingRuntimeManager,
+    set_global_runtime_manager,
+    clear_global_runtime_manager,
+)
+from backend.services.gateway import Gateway
+from backend.services.market import MarketService
+from backend.services.storage import StorageService
+from backend.utils.settlement import SettlementCoordinator
+
+logger = logging.getLogger(__name__)
+_prompt_loader = get_prompt_loader()
+
+
+async def run_gateway(
+    run_id: str,
+    run_dir: Path,
+    bootstrap: dict,
+    port: int
+):
+    """Run Gateway with Pipeline."""
+
+    # Extract config
+    tickers = bootstrap.get("tickers", ["AAPL", "MSFT"])
+    initial_cash = float(bootstrap.get("initial_cash", 100000.0))
+    margin_requirement = float(bootstrap.get("margin_requirement", 0.0))
+    max_comm_cycles = int(bootstrap.get("max_comm_cycles", 2))
+    schedule_mode = bootstrap.get("schedule_mode", "daily")
+    trigger_time = bootstrap.get("trigger_time", "09:30")
+    interval_minutes = int(bootstrap.get("interval_minutes", 60))
+    heartbeat_interval = int(bootstrap.get("heartbeat_interval", 0))  # 0 = disabled
+    mode = bootstrap.get("mode", "live")
+    start_date = bootstrap.get("start_date")
+    end_date = bootstrap.get("end_date")
+    enable_memory = bootstrap.get("enable_memory", False)
+    poll_interval = int(bootstrap.get("poll_interval", 10))
+    enable_mock = bootstrap.get("enable_mock", False)
+
+    is_backtest = mode == "backtest"
+    is_mock = enable_mock or mode == "mock" or (not is_backtest and os.getenv("MOCK_MODE", "false").lower() == "true")
+
+    logger.info(f"[Gateway Server] Starting run {run_id} on port {port}")
+
+    # Create runtime manager
+    runtime_manager = TradingRuntimeManager(
+        config_name=run_id,
+        run_dir=run_dir,
+        bootstrap=bootstrap,
+    )
+    runtime_manager.prepare_run()
+    set_global_runtime_manager(runtime_manager)
+
+    try:
+        async with AsyncExitStack() as stack:
+            # Create services
+            market_service = MarketService(
+                tickers=tickers,
+                poll_interval=poll_interval,
+                mock_mode=is_mock and not is_backtest,
+                backtest_mode=is_backtest,
+                api_key=os.getenv("FINNHUB_API_KEY") if not is_mock and not is_backtest else None,
+                backtest_start_date=start_date if is_backtest else None,
+                backtest_end_date=end_date if is_backtest else None,
+            )
+
+            storage_service = StorageService(
+                dashboard_dir=run_dir / "team_dashboard",
+                initial_cash=initial_cash,
+                config_name=run_id,
+            )
+
+            if not storage_service.files["summary"].exists():
+                storage_service.initialize_empty_dashboard()
+            else:
+                storage_service.update_leaderboard_model_info()
+
+            # Create agents
+            analysts, risk_manager, pm, long_term_memories = create_agents(
+                run_id=run_id,
+                run_dir=run_dir,
+                initial_cash=initial_cash,
+                margin_requirement=margin_requirement,
+                enable_long_term_memory=enable_memory,
+            )
+
+            # Register agents
+            for agent in analysts + [risk_manager, pm]:
+                agent_id = getattr(agent, "agent_id", None) or getattr(agent, "name", None)
+                if agent_id:
+                    runtime_manager.register_agent(agent_id)
+
+            # Load portfolio state
+            portfolio_state = storage_service.load_portfolio_state()
+            pm.load_portfolio_state(portfolio_state)
+
+            # Create settlement coordinator
+            settlement_coordinator = SettlementCoordinator(
+                storage=storage_service,
+                initial_capital=initial_cash,
+            )
+
+            # Create pipeline
+            pipeline = TradingPipeline(
+                analysts=analysts,
+                risk_manager=risk_manager,
+                portfolio_manager=pm,
+                settlement_coordinator=settlement_coordinator,
+                max_comm_cycles=max_comm_cycles,
+                runtime_manager=runtime_manager,
+            )
+
+            # Create scheduler
+            scheduler_callback = None
+            live_scheduler = None
+
+            if is_backtest:
+                backtest_scheduler = BacktestScheduler(
+                    start_date=start_date,
+                    end_date=end_date,
+                    trading_calendar="NYSE",
+                    delay_between_days=0.5,
+                )
+
+                async def scheduler_callback_fn(callback):
+                    await backtest_scheduler.start(callback)
+
+                scheduler_callback = scheduler_callback_fn
+            else:
+                live_scheduler = Scheduler(
+                    mode=schedule_mode,
+                    trigger_time=trigger_time,
+                    interval_minutes=interval_minutes,
+                    heartbeat_interval=heartbeat_interval if heartbeat_interval > 0 else None,
+                    config={"config_name": run_id},
+                )
+
+                async def scheduler_callback_fn(callback):
+                    await live_scheduler.start(callback)
+
+                scheduler_callback = scheduler_callback_fn
+
+            # Enter long-term memory contexts
+            for memory in long_term_memories:
+                await stack.enter_async_context(memory)
+
+            # Create Gateway
+            gateway = Gateway(
+                market_service=market_service,
+                storage_service=storage_service,
+                pipeline=pipeline,
+                scheduler_callback=scheduler_callback,
+                config={
+                    "mode": mode,
+                    "mock_mode": is_mock,
+                    "backtest_mode": is_backtest,
+                    "tickers": tickers,
+                    "config_name": run_id,
+                    "schedule_mode": schedule_mode,
+                    "interval_minutes": interval_minutes,
+                    "trigger_time": trigger_time,
+                    "heartbeat_interval": heartbeat_interval,
+                    "initial_cash": initial_cash,
+                    "margin_requirement": margin_requirement,
+                    "max_comm_cycles": max_comm_cycles,
+                    "enable_memory": enable_memory,
+                },
+                scheduler=live_scheduler,
+            )
+
+            # Start Gateway (blocks until shutdown)
+            logger.info(f"[Gateway Server] Gateway starting on port {port}")
+            await gateway.start(host="0.0.0.0", port=port)
+
+    except asyncio.CancelledError:
+        logger.info("[Gateway Server] Cancelled")
+        raise
+    finally:
+        logger.info("[Gateway Server] Cleaning up")
+        clear_global_runtime_manager()
+
+
+def main():
+    """Main entry point."""
+    parser = argparse.ArgumentParser(description="Gateway Server")
+    parser.add_argument("--run-id", required=True, help="Run identifier")
+    parser.add_argument("--run-dir", required=True, help="Run directory path")
+    parser.add_argument("--port", type=int, default=8765, help="WebSocket port")
+    parser.add_argument("--bootstrap", required=True, help="Bootstrap config as JSON")
+    parser.add_argument("--verbose", action="store_true", help="Verbose logging")
+
+    args = parser.parse_args()
+
+    # Setup logging
+    level = logging.DEBUG if args.verbose else logging.INFO
+    logging.basicConfig(
+        level=level,
+        format="%(asctime)s | %(levelname)-7s | %(name)s:%(lineno)d - %(message)s",
+    )
+
+    # Parse bootstrap
+    bootstrap = json.loads(args.bootstrap)
+    run_dir = Path(args.run_dir)
+
+    # Run
+    try:
+        asyncio.run(run_gateway(
+            run_id=args.run_id,
+            run_dir=run_dir,
+            bootstrap=bootstrap,
+            port=args.port
+        ))
+    except KeyboardInterrupt:
+        logger.info("[Gateway Server] Interrupted by user")
+    except Exception as e:
+        logger.exception(f"[Gateway Server] Fatal error: {e}")
+        sys.exit(1)
+
+
+if __name__ == "__main__":
+    main()
--- a/backend/llm/models.py
+++ b/backend/llm/models.py
@@ -3,9 +3,11 @@
 AgentScope Native Model Factory
 Uses native AgentScope model classes for LLM calls
 """
-from enum import Enum
-from typing import Optional, Tuple
 import os
+import time
+import logging
+from enum import Enum
+from typing import Any, Callable, Optional, Tuple, TypeVar, Union
 from agentscope.formatter import (
    AnthropicChatFormatter,
    DashScopeChatFormatter,
@@ -26,6 +28,244 @@ from backend.config.env_config import (
    get_env_str,
 )

+logger = logging.getLogger(__name__)
+
+# Retry wrapper types
+T = TypeVar("T")
+
+
+class RetryChatModel:
+    """Wraps an AgentScope model with automatic retry for transient errors.
+
+    Based on CoPaw's RetryChatModel design. Handles rate limits, timeouts,
+    and other transient failures with exponential backoff.
+    """
+
+    DEFAULT_MAX_RETRIES = 3
+    DEFAULT_INITIAL_DELAY = 1.0
+    DEFAULT_MAX_DELAY = 60.0
+    DEFAULT_BACKOFF_MULTIPLIER = 2.0
+
+    # Transient error codes/messages that should trigger retry
+    TRANSIENT_ERROR_KEYWORDS = frozenset([
+        "rate_limit",
+        "429",
+        "timeout",
+        "503",
+        "502",
+        "504",
+        "connection",
+        "temporary",
+        "overloaded",
+        "too_many_requests",
+    ])
+
+    def __init__(
+        self,
+        model: Any,
+        max_retries: int = DEFAULT_MAX_RETRIES,
+        initial_delay: float = DEFAULT_INITIAL_DELAY,
+        max_delay: float = DEFAULT_MAX_DELAY,
+        backoff_multiplier: float = DEFAULT_BACKOFF_MULTIPLIER,
+        on_retry: Optional[Callable[[int, Exception, float], None]] = None,
+    ):
+        """Initialize retry wrapper.
+
+        Args:
+            model: The underlying AgentScope model to wrap
+            max_retries: Maximum number of retry attempts
+            initial_delay: Initial delay in seconds before first retry
+            max_delay: Maximum delay between retries
+            backoff_multiplier: Multiplier for exponential backoff
+            on_retry: Optional callback(retry_count, exception, delay) for logging
+        """
+        self._model = model
+        self._max_retries = max_retries
+        self._initial_delay = initial_delay
+        self._max_delay = max_delay
+        self._backoff_multiplier = backoff_multiplier
+        self._on_retry = on_retry
+        self._total_tokens_used = 0
+        self._total_cost = 0.0
+
+    @property
+    def model_name(self) -> str:
+        return getattr(self._model, "model_name", str(self._model))
+
+    @property
+    def total_tokens_used(self) -> int:
+        return self._total_tokens_used
+
+    @property
+    def total_cost(self) -> float:
+        return self._total_cost
+
+    def _is_transient_error(self, error: Exception) -> bool:
+        """Check if an error is transient and should be retried.
+
+        Args:
+            error: The exception to check
+
+        Returns:
+            True if the error is transient
+        """
+        error_str = str(error).lower()
+        for keyword in self.TRANSIENT_ERROR_KEYWORDS:
+            if keyword in error_str:
+                return True
+        return False
+
+    def _calculate_delay(self, retry_count: int) -> float:
+        """Calculate delay for given retry attempt with exponential backoff.
+
+        Args:
+            retry_count: Current retry attempt number (1-based)
+
+        Returns:
+            Delay in seconds
+        """
+        delay = self._initial_delay * (self._backoff_multiplier ** (retry_count - 1))
+        return min(delay, self._max_delay)
+
+    def _call_with_retry(self, func: Callable[..., T], *args, **kwargs) -> T:
+        """Call a function with retry logic for transient errors.
+
+        Args:
+            func: Function to call
+            *args: Positional arguments
+            **kwargs: Keyword arguments
+
+        Returns:
+            Result from func
+
+        Raises:
+            Last exception if all retries exhausted
+        """
+        last_error: Optional[Exception] = None
+
+        for attempt in range(1, self._max_retries + 1):
+            try:
+                result = func(*args, **kwargs)
+
+                # Track usage if available
+                if hasattr(result, "usage") and result.usage:
+                    usage = result.usage
+                    self._total_tokens_used += getattr(usage, "total_tokens", 0)
+                    self._total_cost += getattr(usage, "cost", 0.0)
+
+                return result
+
+            except Exception as e:
+                last_error = e
+
+                if attempt >= self._max_retries:
+                    logger.error(
+                        "RetryChatModel: Max retries (%d) exhausted for %s",
+                        self._max_retries,
+                        self.model_name,
+                    )
+                    break
+
+                if not self._is_transient_error(e):
+                    logger.warning(
+                        "RetryChatModel: Non-transient error, not retrying: %s",
+                        str(e),
+                    )
+                    break
+
+                delay = self._calculate_delay(attempt)
+                logger.warning(
+                    "RetryChatModel: Transient error on attempt %d/%d, "
+                    "retrying in %.1fs: %s",
+                    attempt,
+                    self._max_retries,
+                    delay,
+                    str(e)[:200],
+                )
+
+                if self._on_retry:
+                    self._on_retry(attempt, e, delay)
+
+                time.sleep(delay)
+
+        if last_error is not None:
+            raise last_error
+        raise RuntimeError("RetryChatModel: Unexpected state, no error but no result")
+
+    def __call__(self, *args, **kwargs) -> Any:
+        """Forward calls to the wrapped model with retry logic."""
+        return self._call_with_retry(self._model, *args, **kwargs)
+
+    def __getattr__(self, name: str) -> Any:
+        """Proxy attribute access to the wrapped model."""
+        return getattr(self._model, name)
+
+
+class TokenRecordingModelWrapper:
+    """Wraps a model to track token usage per provider.
+
+    Based on CoPaw's TokenRecordingModelWrapper design.
+    """
+
+    def __init__(self, model: Any):
+        """Initialize token recorder.
+
+        Args:
+            model: The underlying AgentScope model to wrap
+        """
+        self._model = model
+        self._total_tokens = 0
+        self._prompt_tokens = 0
+        self._completion_tokens = 0
+        self._total_cost = 0.0
+
+    @property
+    def model_name(self) -> str:
+        return getattr(self._model, "model_name", str(self._model))
+
+    @property
+    def total_tokens(self) -> int:
+        return self._total_tokens
+
+    @property
+    def prompt_tokens(self) -> int:
+        return self._prompt_tokens
+
+    @property
+    def completion_tokens(self) -> int:
+        return self._completion_tokens
+
+    @property
+    def total_cost(self) -> float:
+        return self._total_cost
+
+    def record_usage(self, usage: Any) -> None:
+        """Record token usage from a model response.
+
+        Args:
+            usage: Usage object from model response
+        """
+        if usage is None:
+            return
+
+        self._prompt_tokens += getattr(usage, "prompt_tokens", 0)
+        self._completion_tokens += getattr(usage, "completion_tokens", 0)
+        self._total_tokens += getattr(usage, "total_tokens", 0)
+        self._total_cost += getattr(usage, "cost", 0.0)
+
+    def __call__(self, *args, **kwargs) -> Any:
+        """Forward calls and record usage."""
+        result = self._model(*args, **kwargs)
+
+        if hasattr(result, "usage") and result.usage:
+            self.record_usage(result.usage)
+
+        return result
+
+    def __getattr__(self, name: str) -> Any:
+        """Proxy attribute access to the wrapped model."""
+        return getattr(self._model, name)
+

 class ModelProvider(Enum):
    """Supported model providers"""
--- a/backend/main.py
+++ b/backend/main.py
@@ -14,20 +14,50 @@ import loguru
 from dotenv import load_dotenv

 from backend.agents import AnalystAgent, PMAgent, RiskAgent
+from backend.agents.skills_manager import SkillsManager
+from backend.agents.toolkit_factory import create_agent_toolkit, load_agent_profiles
+from backend.agents.prompt_loader import get_prompt_loader
+from backend.agents.workspace_manager import WorkspaceManager
+from backend.config.bootstrap_config import resolve_runtime_config
 from backend.config.constants import ANALYST_TYPES
-from backend.config.env_config import get_env_float, get_env_int, get_env_list
 from backend.core.pipeline import TradingPipeline
 from backend.core.scheduler import BacktestScheduler, Scheduler
-from backend.utils.settlement import SettlementCoordinator
 from backend.llm.models import get_agent_formatter, get_agent_model
+from backend.api.runtime import register_runtime_manager, unregister_runtime_manager
+from backend.runtime.manager import (
+    TradingRuntimeManager,
+    set_global_runtime_manager,
+    clear_global_runtime_manager,
+)
 from backend.services.gateway import Gateway
 from backend.services.market import MarketService
 from backend.services.storage import StorageService
+from backend.utils.settlement import SettlementCoordinator

 load_dotenv()
 logger = logging.getLogger(__name__)
 loguru.logger.disable("flowllm")
 loguru.logger.disable("reme_ai")
+_prompt_loader = get_prompt_loader()
+
+
+def _get_run_dir(config_name: str) -> Path:
+    """Return the canonical run-scoped directory for a config."""
+    project_root = Path(__file__).resolve().parents[1]
+    return WorkspaceManager(project_root=project_root).get_run_dir(config_name)
+
+
+def _resolve_runtime_config(args) -> dict:
+    """Merge env defaults with run-scoped bootstrap config."""
+    project_root = Path(__file__).resolve().parents[1]
+    return resolve_runtime_config(
+        project_root=project_root,
+        config_name=args.config_name,
+        enable_memory=args.enable_memory,
+        schedule_mode=args.schedule_mode,
+        interval_minutes=args.interval_minutes,
+        trigger_time=args.trigger_time,
+    )


 def create_long_term_memory(agent_name: str, config_name: str):
@@ -45,7 +75,7 @@ def create_long_term_memory(agent_name: str, config_name: str):
        logger.warning("MEMORY_API_KEY not set, long-term memory disabled")
        return None

-    memory_dir = str(Path(config_name) / "memory")
+    memory_dir = str(_get_run_dir(config_name) / "memory")

    return ReMeTaskLongTermMemory(
        agent_name=agent_name,
@@ -84,11 +114,31 @@ def create_agents(
    """
    analysts = []
    long_term_memories = []
+    workspace_manager = WorkspaceManager()
+    workspace_manager.initialize_default_assets(
+        config_name=config_name,
+        agent_ids=list(ANALYST_TYPES.keys())
+        + ["risk_manager", "portfolio_manager"],
+        analyst_personas=_prompt_loader.load_yaml_config("analyst", "personas"),
+    )
+    profiles = load_agent_profiles()
+    skills_manager = SkillsManager()
+    active_skill_map = skills_manager.prepare_active_skills(
+        config_name=config_name,
+        agent_defaults={
+            agent_id: profile.get("skills", [])
+            for agent_id, profile in profiles.items()
+        },
+    )

    for analyst_type in ANALYST_TYPES:
        model = get_agent_model(analyst_type)
        formatter = get_agent_formatter(analyst_type)
-        toolkit = create_toolkit(analyst_type)
+        toolkit = create_agent_toolkit(
+            analyst_type,
+            config_name,
+            active_skill_dirs=active_skill_map.get(analyst_type, []),
+        )

        long_term_memory = None
        if enable_long_term_memory:
@@ -125,6 +175,11 @@ def create_agents(
        name="risk_manager",
        config={"config_name": config_name},
        long_term_memory=risk_long_term_memory,
+        toolkit=create_agent_toolkit(
+            "risk_manager",
+            config_name,
+            active_skill_dirs=active_skill_map.get("risk_manager", []),
+        ),
    )

    pm_long_term_memory = None
@@ -144,44 +199,34 @@ def create_agents(
        margin_requirement=margin_requirement,
        config={"config_name": config_name},
        long_term_memory=pm_long_term_memory,
+        toolkit_factory=create_agent_toolkit,
+        toolkit_factory_kwargs={
+            "active_skill_dirs": active_skill_map.get(
+                "portfolio_manager",
+                [],
+            ),
+        },
    )

    return analysts, risk_manager, portfolio_manager, long_term_memories
-
-
-def create_toolkit(analyst_type: str):
-    """Create AgentScope Toolkit with tools for specific analyst type"""
-    from agentscope.tool import Toolkit
-    from backend.agents.prompt_loader import PromptLoader
-    from backend.tools.analysis_tools import TOOL_REGISTRY
-
-    # Load analyst persona config
-    prompt_loader = PromptLoader()
-    personas_config = prompt_loader.load_yaml_config("analyst", "personas")
-    persona = personas_config.get(analyst_type, {})
-
-    # Get tool names for this analyst type
-    tool_names = persona.get("tools", [])
-
-    # Create toolkit and register tools
-    toolkit = Toolkit()
-    for tool_name in tool_names:
-        tool_func = TOOL_REGISTRY.get(tool_name)
-        if tool_func:
-            toolkit.register_tool_function(tool_func)
-
-    return toolkit
-
-
 async def run_with_gateway(args):
    """Run with WebSocket gateway"""
    is_backtest = args.mode == "backtest"
+    runtime_config = _resolve_runtime_config(args)

-    # Load config from env, override with args
-    tickers = get_env_list("TICKERS", ["AAPL", "MSFT"])
-    initial_cash = get_env_float("INITIAL_CASH", 100000.0)
-    margin_requirement = get_env_float("MARGIN_REQUIREMENT", 0.0)
    config_name = args.config_name
+    tickers = runtime_config["tickers"]
+    initial_cash = runtime_config["initial_cash"]
+    margin_requirement = runtime_config["margin_requirement"]
+
+    runtime_manager = TradingRuntimeManager(
+        config_name=config_name,
+        run_dir=_get_run_dir(config_name),
+        bootstrap=runtime_config,
+    )
+    runtime_manager.prepare_run()
+    set_global_runtime_manager(runtime_manager)
+    register_runtime_manager(runtime_manager)

    # Create market service
    market_service = MarketService(
@@ -198,7 +243,7 @@ async def run_with_gateway(args):

    # Create storage service
    storage_service = StorageService(
-        dashboard_dir=Path(config_name) / "team_dashboard",
+        dashboard_dir=_get_run_dir(config_name) / "team_dashboard",
        initial_cash=initial_cash,
        config_name=config_name,
    )
@@ -213,8 +258,12 @@ async def run_with_gateway(args):
        config_name=config_name,
        initial_cash=initial_cash,
        margin_requirement=margin_requirement,
-        enable_long_term_memory=args.enable_memory,
+        enable_long_term_memory=runtime_config["enable_memory"],
    )
+    for agent in analysts + [risk_manager, pm]:
+        agent_id = getattr(agent, "agent_id", None) or getattr(agent, "name", None)
+        if agent_id:
+            runtime_manager.register_agent(agent_id)
    portfolio_state = storage_service.load_portfolio_state()
    pm.load_portfolio_state(portfolio_state)

@@ -228,12 +277,14 @@ async def run_with_gateway(args):
        risk_manager=risk_manager,
        portfolio_manager=pm,
        settlement_coordinator=settlement_coordinator,
-        max_comm_cycles=get_env_int("MAX_COMM_CYCLES", 2),
+        max_comm_cycles=runtime_config["max_comm_cycles"],
+        runtime_manager=runtime_manager,
    )

    # Create scheduler callback
    scheduler_callback = None
    trading_dates = []
+    live_scheduler = None

    if is_backtest:
        backtest_scheduler = BacktestScheduler(
@@ -249,10 +300,11 @@ async def run_with_gateway(args):

        scheduler_callback = scheduler_callback_fn
    else:
-        # Live mode: use daily scheduler with NYSE timezone
+        # Live mode: use daily or intraday scheduler with NYSE timezone
        live_scheduler = Scheduler(
-            mode="daily",
-            trigger_time=args.trigger_time,
+            mode=runtime_config["schedule_mode"],
+            trigger_time=runtime_config["trigger_time"],
+            interval_minutes=runtime_config["interval_minutes"],
            config={"config_name": config_name},
        )

@@ -273,7 +325,15 @@ async def run_with_gateway(args):
            "backtest_mode": is_backtest,
            "tickers": tickers,
            "config_name": config_name,
+            "schedule_mode": runtime_config["schedule_mode"],
+            "interval_minutes": runtime_config["interval_minutes"],
+            "trigger_time": runtime_config["trigger_time"],
+            "initial_cash": initial_cash,
+            "margin_requirement": margin_requirement,
+            "max_comm_cycles": runtime_config["max_comm_cycles"],
+            "enable_memory": runtime_config["enable_memory"],
        },
+        scheduler=live_scheduler if not is_backtest else None,
    )

    if is_backtest:
@@ -281,9 +341,13 @@ async def run_with_gateway(args):

    # Start long-term memory contexts and run gateway
    async with AsyncExitStack() as stack:
-        for memory in long_term_memories:
-            await stack.enter_async_context(memory)
-        await gateway.start(host=args.host, port=args.port)
+        try:
+            for memory in long_term_memories:
+                await stack.enter_async_context(memory)
+            await gateway.start(host=args.host, port=args.port)
+        finally:
+            unregister_runtime_manager()
+            clear_global_runtime_manager()


 def main():
@@ -294,7 +358,13 @@ def main():
    parser.add_argument("--config-name", default="mock")
    parser.add_argument("--host", default="0.0.0.0")
    parser.add_argument("--port", type=int, default=8765)
+    parser.add_argument(
+        "--schedule-mode",
+        choices=["daily", "intraday"],
+        default="daily",
+    )
    parser.add_argument("--trigger-time", default="09:30")  # NYSE market open
+    parser.add_argument("--interval-minutes", type=int, default=60)
    parser.add_argument("--poll-interval", type=int, default=10)
    parser.add_argument("--start-date")
    parser.add_argument("--end-date")
@@ -307,15 +377,17 @@ def main():
    args = parser.parse_args()

    # Load config from env for logging
-    tickers = get_env_list("TICKERS", ["AAPL", "MSFT"])
-    initial_cash = get_env_float("INITIAL_CASH", 100000.0)
+    runtime_config = _resolve_runtime_config(args)
+    tickers = runtime_config["tickers"]
+    initial_cash = runtime_config["initial_cash"]

    logger.info("=" * 60)
    logger.info(f"Mode: {args.mode}, Config: {args.config_name}")
    logger.info(f"Tickers: {tickers}")
    logger.info(f"Initial Cash: ${initial_cash:,.2f}")
    logger.info(
-        f"Long-term Memory: {'enabled' if args.enable_memory else 'disabled'}",
+        "Long-term Memory: %s",
+        "enabled" if runtime_config["enable_memory"] else "disabled",
    )
    if args.mode == "backtest":
        if not args.start_date or not args.end_date:
--- a/backend/process/models.py
+++ b/backend/process/models.py
@@ -0,0 +1,41 @@
+# -*- coding: utf-8 -*-
+"""Data models for lightweight process supervision."""
+
+from dataclasses import dataclass, field
+from datetime import datetime
+from enum import Enum
+from typing import Any, Dict
+
+
+class ProcessRunState(str, Enum):
+    """Execution state for supervised runs."""
+
+    PENDING = "pending"
+    RUNNING = "running"
+    COMPLETED = "completed"
+    FAILED = "failed"
+    CANCELLED = "cancelled"
+
+
+@dataclass
+class ProcessRun:
+    """Represents a supervised process run."""
+
+    run_id: str
+    command: str
+    scope_key: str
+    state: ProcessRunState = ProcessRunState.PENDING
+    metadata: Dict[str, Any] = field(default_factory=dict)
+    created_at: datetime = field(default_factory=datetime.utcnow)
+    updated_at: datetime = field(default_factory=datetime.utcnow)
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "run_id": self.run_id,
+            "command": self.command,
+            "scope_key": self.scope_key,
+            "state": self.state.value,
+            "metadata": self.metadata,
+            "created_at": self.created_at.isoformat(),
+            "updated_at": self.updated_at.isoformat(),
+        }
--- a/backend/process/registry.py
+++ b/backend/process/registry.py
@@ -0,0 +1,35 @@
+# -*- coding: utf-8 -*-
+"""Registry for managing supervised process metadata."""
+
+from threading import Lock
+from typing import Dict, Iterable, Optional
+
+from .models import ProcessRun
+
+
+class RunRegistry:
+    """In-memory registry for tracked process runs."""
+
+    def __init__(self) -> None:
+        self._runs: Dict[str, ProcessRun] = {}
+        self._lock = Lock()
+
+    def add(self, run: ProcessRun) -> None:
+        with self._lock:
+            self._runs[run.run_id] = run
+
+    def get(self, run_id: str) -> Optional[ProcessRun]:
+        with self._lock:
+            return self._runs.get(run_id)
+
+    def list(self) -> Iterable[ProcessRun]:
+        with self._lock:
+            return list(self._runs.values())
+
+    def update(self, run: ProcessRun) -> None:
+        with self._lock:
+            self._runs[run.run_id] = run
+
+    def remove(self, run_id: str) -> None:
+        with self._lock:
+            self._runs.pop(run_id, None)
--- a/backend/process/supervisor.py
+++ b/backend/process/supervisor.py
@@ -0,0 +1,61 @@
+# -*- coding: utf-8 -*-
+"""Minimal supervisor for scripted tasks and long-running utilities."""
+
+from datetime import datetime
+from typing import Any, Dict, Iterable, Optional
+
+from .models import ProcessRun, ProcessRunState
+from .registry import RunRegistry
+
+
+class ProcessSupervisor:
+    """Tracks supervised runs without executing real processes yet."""
+
+    def __init__(self, registry: Optional[RunRegistry] = None) -> None:
+        self.registry = registry or RunRegistry()
+
+    def spawn(
+        self,
+        run_id: str,
+        command: str,
+        scope_key: str,
+        metadata: Optional[Dict[str, Any]] = None,
+    ) -> ProcessRun:
+        run = ProcessRun(
+            run_id=run_id,
+            command=command,
+            scope_key=scope_key,
+            metadata=metadata or {},
+        )
+        run.state = ProcessRunState.RUNNING
+        run.updated_at = datetime.utcnow()
+        self.registry.add(run)
+        return run
+
+    def update_state(
+        self,
+        run_id: str,
+        state: ProcessRunState,
+        metadata: Optional[Dict[str, Any]] = None,
+    ) -> Optional[ProcessRun]:
+        run = self.registry.get(run_id)
+        if not run:
+            return None
+        run.state = state
+        run.metadata.update(metadata or {})
+        run.updated_at = datetime.utcnow()
+        self.registry.update(run)
+        return run
+
+    def cancel(self, run_id: str, reason: Optional[str] = None) -> Optional[ProcessRun]:
+        run = self.registry.get(run_id)
+        if not run:
+            return None
+        run.state = ProcessRunState.CANCELLED
+        run.metadata.setdefault("cancel_reason", reason or "manual")
+        run.updated_at = datetime.utcnow()
+        self.registry.update(run)
+        return run
+
+    def list_runs(self) -> Iterable[ProcessRun]:
+        return self.registry.list()
--- a/backend/runtime/init.py
+++ b/backend/runtime/init.py
@@ -0,0 +1,13 @@
+from .agent_runtime import AgentRuntimeState
+from .context import TradingRunContext
+from .manager import TradingRuntimeManager
+from .registry import RuntimeRegistry
+from .session import TradingSessionKey
+
+__all__ = [
+    "AgentRuntimeState",
+    "TradingRunContext",
+    "TradingRuntimeManager",
+    "RuntimeRegistry",
+    "TradingSessionKey",
+]
--- a/backend/runtime/agent_runtime.py
+++ b/backend/runtime/agent_runtime.py
@@ -0,0 +1,26 @@
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from datetime import datetime, UTC
+from typing import Any, Dict
+
+
+@dataclass
+class AgentRuntimeState:
+    agent_id: str
+    status: str = "idle"
+    last_session: str | None = None
+    last_updated: datetime = field(default_factory=lambda: datetime.now(UTC))
+
+    def update(self, status: str, session_key: str | None = None) -> None:
+        self.status = status
+        self.last_session = session_key
+        self.last_updated = datetime.now(UTC)
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "agent_id": self.agent_id,
+            "status": self.status,
+            "last_session": self.last_session,
+            "last_updated": self.last_updated.isoformat(),
+        }
--- a/backend/runtime/context.py
+++ b/backend/runtime/context.py
@@ -0,0 +1,15 @@
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Dict
+
+
+@dataclass(frozen=True)
+class TradingRunContext:
+    config_name: str
+    run_dir: Path
+    bootstrap_values: Dict[str, Any] = field(default_factory=dict)
+
+    def describe(self) -> str:
+        return f"Run {self.config_name} @ {self.run_dir}"
--- a/backend/runtime/manager.py
+++ b/backend/runtime/manager.py
@@ -0,0 +1,158 @@
+from __future__ import annotations
+
+import asyncio
+import json
+from datetime import datetime, UTC
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+from .agent_runtime import AgentRuntimeState
+from .context import TradingRunContext
+from .registry import RuntimeRegistry
+
+_global_runtime_manager: Optional["TradingRuntimeManager"] = None
+_shutdown_event: Optional[asyncio.Event] = None
+
+
+def set_global_runtime_manager(manager: "TradingRuntimeManager") -> None:
+    global _global_runtime_manager
+    _global_runtime_manager = manager
+
+
+def clear_global_runtime_manager() -> None:
+    global _global_runtime_manager
+    _global_runtime_manager = None
+
+
+def get_global_runtime_manager() -> Optional["TradingRuntimeManager"]:
+    return _global_runtime_manager
+
+
+def set_shutdown_event(event: asyncio.Event) -> None:
+    """Set the global shutdown event for signaling runtime stop."""
+    global _shutdown_event
+    _shutdown_event = event
+
+
+def clear_shutdown_event() -> None:
+    """Clear the global shutdown event."""
+    global _shutdown_event
+    _shutdown_event = None
+
+
+def get_shutdown_event() -> Optional[asyncio.Event]:
+    """Get the global shutdown event if set."""
+    return _shutdown_event
+
+
+def is_shutdown_requested() -> bool:
+    """Check if shutdown has been requested."""
+    return _shutdown_event is not None and _shutdown_event.is_set()
+
+
+class TradingRuntimeManager:
+    def __init__(self, config_name: str, run_dir: Path, bootstrap: Optional[Dict[str, Any]] = None) -> None:
+        self.config_name = config_name
+        self.run_dir = run_dir
+        self.bootstrap = bootstrap or {}
+        self.context: Optional[TradingRunContext] = None
+        self.registry = RuntimeRegistry()
+        self.current_session_key: Optional[str] = None
+        self.events: List[Dict[str, Any]] = []
+        self.pending_approvals: Dict[str, Dict[str, Any]] = {}
+        self.snapshot_path = self.run_dir / "state" / "runtime_state.json"
+
+    def prepare_run(self) -> TradingRunContext:
+        self.run_dir.mkdir(parents=True, exist_ok=True)
+        self.context = TradingRunContext(
+            config_name=self.config_name,
+            run_dir=self.run_dir,
+            bootstrap_values=self.bootstrap,
+        )
+        self._persist_snapshot()
+        return self.context
+
+    def set_session_key(self, session_key: str) -> None:
+        self.current_session_key = session_key
+        self._persist_snapshot()
+
+    def log_event(self, event: str, details: Optional[Dict[str, Any]] = None) -> Dict[str, Any]:
+        entry = {
+            "timestamp": datetime.now(UTC).isoformat(),
+            "event": event,
+            "details": details or {},
+            "session": self.current_session_key,
+        }
+        self.events.append(entry)
+        self._persist_snapshot()
+        return entry
+
+    def register_agent(self, agent_id: str) -> AgentRuntimeState:
+        state = AgentRuntimeState(agent_id=agent_id)
+        self.registry.register(agent_id, state)
+        self._persist_snapshot()
+        return state
+
+    def register_pending_approval(self, approval_id: str, payload: Dict[str, Any]) -> None:
+        payload.setdefault("status", "pending")
+        payload.setdefault("created_at", datetime.now(UTC).isoformat())
+        self.pending_approvals[approval_id] = payload
+        self._persist_snapshot()
+
+    def update_agent_status(
+        self,
+        agent_id: str,
+        status: str,
+        session_key: Optional[str] = None,
+    ) -> AgentRuntimeState:
+        state = self.registry.get(agent_id)
+        if state is None:
+            state = self.register_agent(agent_id)
+        effective_session = session_key or self.current_session_key
+        state.update(status, effective_session)
+        self._persist_snapshot()
+        return state
+
+    def get_agent_state(self, agent_id: str) -> Optional[AgentRuntimeState]:
+        return self.registry.get(agent_id)
+
+    def list_agents(self) -> list[str]:
+        return self.registry.list_agents()
+
+    def resolve_pending_approval(self, approval_id: str, resolved_by: str, status: str) -> None:
+        entry = self.pending_approvals.get(approval_id)
+        if not entry:
+            return
+        entry["status"] = status
+        entry["resolved_at"] = datetime.now(UTC).isoformat()
+        entry["resolved_by"] = resolved_by
+        self._persist_snapshot()
+
+    def list_pending_approvals(self) -> List[Dict[str, Any]]:
+        return list(self.pending_approvals.values())
+
+    def build_snapshot(self) -> Dict[str, Any]:
+        return {
+            "context": {
+                "config_name": self.context.config_name,
+                "run_dir": str(self.context.run_dir),
+                "bootstrap_values": self.context.bootstrap_values,
+            }
+            if self.context
+            else None,
+            "current_session_key": self.current_session_key,
+            "agents": [
+                state.to_dict()
+                for agent_id in self.registry.list_agents()
+                if (state := self.registry.get(agent_id)) is not None
+            ],
+            "events": self.events,
+            "pending_approvals": self.list_pending_approvals(),
+        }
+
+    def _persist_snapshot(self) -> None:
+        self.snapshot_path.parent.mkdir(parents=True, exist_ok=True)
+        self.snapshot_path.write_text(
+            json.dumps(self.build_snapshot(), ensure_ascii=False, indent=2),
+            encoding="utf-8",
+        )
--- a/backend/runtime/registry.py
+++ b/backend/runtime/registry.py
@@ -0,0 +1,20 @@
+from __future__ import annotations
+
+from typing import Dict, Optional
+
+
+class RuntimeRegistry:
+    def __init__(self) -> None:
+        self._states: Dict[str, "AgentRuntimeState"] = {}
+
+    def register(self, agent_id: str, state: "AgentRuntimeState") -> None:
+        self._states[agent_id] = state
+
+    def get(self, agent_id: str) -> Optional["AgentRuntimeState"]:
+        return self._states.get(agent_id)
+
+    def list_agents(self) -> list[str]:
+        return list(self._states.keys())
+
+    def clear(self) -> None:
+        self._states.clear()
--- a/backend/runtime/session.py
+++ b/backend/runtime/session.py
@@ -0,0 +1,14 @@
+from dataclasses import dataclass
+
+
+@dataclass(frozen=True)
+class TradingSessionKey:
+    date: str
+    ticker: str | None = None
+
+    def __post_init__(self):
+        if not self.date:
+            raise ValueError("Session must have a date")
+
+    def key(self) -> str:
+        return f"{self.date}:{self.ticker or 'all'}"
--- a/backend/services/gateway.py
+++ b/backend/services/gateway.py
--- a/backend/services/gateway_admin_handlers.py
+++ b/backend/services/gateway_admin_handlers.py
@@ -0,0 +1,419 @@
+# -*- coding: utf-8 -*-
+"""Runtime/workspace/skills handlers extracted from the main Gateway module."""
+
+from __future__ import annotations
+
+import json
+from datetime import datetime
+from typing import Any
+
+from backend.agents.agent_workspace import load_agent_workspace_config
+from backend.agents.skills_manager import SkillsManager
+from backend.agents.toolkit_factory import load_agent_profiles
+from backend.config.bootstrap_config import (
+    get_bootstrap_config_for_run,
+    resolve_runtime_config,
+    update_bootstrap_values_for_run,
+)
+from backend.data.market_ingest import ingest_symbols
+from backend.llm.models import get_agent_model_info
+
+
+async def handle_reload_runtime_assets(gateway: Any) -> None:
+    config_name = gateway.config.get("config_name", "default")
+    runtime_config = resolve_runtime_config(
+        project_root=gateway._project_root,
+        config_name=config_name,
+        enable_memory=gateway.config.get("enable_memory", False),
+        schedule_mode=gateway.config.get("schedule_mode", "daily"),
+        interval_minutes=gateway.config.get("interval_minutes", 60),
+        trigger_time=gateway.config.get("trigger_time", "09:30"),
+    )
+    result = gateway.pipeline.reload_runtime_assets(runtime_config=runtime_config)
+    runtime_updates = gateway._apply_runtime_config(runtime_config)
+    await gateway.state_sync.on_system_message("Runtime assets reloaded.")
+    await gateway.broadcast({"type": "runtime_assets_reloaded", **result, **runtime_updates})
+
+
+async def handle_update_runtime_config(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    updates: dict[str, Any] = {}
+
+    schedule_mode = str(data.get("schedule_mode", "")).strip().lower()
+    if schedule_mode:
+        if schedule_mode not in {"daily", "intraday"}:
+            await websocket.send(json.dumps({"type": "error", "message": "schedule_mode must be 'daily' or 'intraday'."}, ensure_ascii=False))
+            return
+        updates["schedule_mode"] = schedule_mode
+
+    interval_minutes = data.get("interval_minutes")
+    if interval_minutes is not None:
+        try:
+            parsed_interval = int(interval_minutes)
+        except (TypeError, ValueError):
+            parsed_interval = 0
+        if parsed_interval <= 0:
+            await websocket.send(json.dumps({"type": "error", "message": "interval_minutes must be a positive integer."}, ensure_ascii=False))
+            return
+        updates["interval_minutes"] = parsed_interval
+
+    trigger_time = data.get("trigger_time")
+    if trigger_time is not None:
+        raw_trigger = str(trigger_time).strip()
+        if raw_trigger and raw_trigger != "now":
+            try:
+                datetime.strptime(raw_trigger, "%H:%M")
+            except ValueError:
+                await websocket.send(json.dumps({"type": "error", "message": "trigger_time must use HH:MM or 'now'."}, ensure_ascii=False))
+                return
+        updates["trigger_time"] = raw_trigger or "09:30"
+
+    max_comm_cycles = data.get("max_comm_cycles")
+    if max_comm_cycles is not None:
+        try:
+            parsed_cycles = int(max_comm_cycles)
+        except (TypeError, ValueError):
+            parsed_cycles = 0
+        if parsed_cycles <= 0:
+            await websocket.send(json.dumps({"type": "error", "message": "max_comm_cycles must be a positive integer."}, ensure_ascii=False))
+            return
+        updates["max_comm_cycles"] = parsed_cycles
+
+    initial_cash = data.get("initial_cash")
+    if initial_cash is not None:
+        try:
+            parsed_initial_cash = float(initial_cash)
+        except (TypeError, ValueError):
+            parsed_initial_cash = 0.0
+        if parsed_initial_cash <= 0:
+            await websocket.send(json.dumps({"type": "error", "message": "initial_cash must be a positive number."}, ensure_ascii=False))
+            return
+        updates["initial_cash"] = parsed_initial_cash
+
+    margin_requirement = data.get("margin_requirement")
+    if margin_requirement is not None:
+        try:
+            parsed_margin_requirement = float(margin_requirement)
+        except (TypeError, ValueError):
+            parsed_margin_requirement = -1.0
+        if parsed_margin_requirement < 0:
+            await websocket.send(json.dumps({"type": "error", "message": "margin_requirement must be a non-negative number."}, ensure_ascii=False))
+            return
+        updates["margin_requirement"] = parsed_margin_requirement
+
+    enable_memory = data.get("enable_memory")
+    if enable_memory is not None:
+        updates["enable_memory"] = bool(enable_memory)
+
+    if not updates:
+        await websocket.send(json.dumps({"type": "error", "message": "No runtime settings were provided."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    update_bootstrap_values_for_run(
+        project_root=gateway._project_root,
+        config_name=config_name,
+        updates=updates,
+    )
+    await gateway.state_sync.on_system_message("运行时调度配置已保存，正在热更新")
+    await handle_reload_runtime_assets(gateway)
+
+
+async def handle_update_watchlist(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    tickers = gateway._normalize_watchlist(data.get("tickers"))
+    if not tickers:
+        await websocket.send(json.dumps({"type": "error", "message": "update_watchlist requires at least one valid ticker."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    update_bootstrap_values_for_run(
+        project_root=gateway._project_root,
+        config_name=config_name,
+        updates={"tickers": tickers},
+    )
+    await gateway.state_sync.on_system_message(f"Watchlist updated: {', '.join(tickers)}")
+    await gateway.broadcast({"type": "watchlist_updated", "config_name": config_name, "tickers": tickers})
+    await handle_reload_runtime_assets(gateway)
+    gateway._schedule_watchlist_market_store_refresh(tickers)
+
+
+async def handle_get_agent_skills(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    if not agent_id:
+        await websocket.send(json.dumps({"type": "error", "message": "get_agent_skills requires agent_id."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    agent_asset_dir = skills_manager.get_agent_asset_dir(config_name, agent_id)
+    agent_config = load_agent_workspace_config(agent_asset_dir / "agent.yaml")
+    resolved_skills = set(skills_manager.resolve_agent_skill_names(config_name=config_name, agent_id=agent_id, default_skills=[]))
+    enabled = set(agent_config.enabled_skills)
+    disabled = set(agent_config.disabled_skills)
+
+    payload = []
+    for item in skills_manager.list_agent_skill_catalog(config_name, agent_id):
+        if item.skill_name in disabled:
+            status = "disabled"
+        elif item.skill_name in enabled:
+            status = "enabled"
+        elif item.skill_name in resolved_skills:
+            status = "active"
+        else:
+            status = "available"
+        payload.append({
+            "skill_name": item.skill_name,
+            "name": item.name,
+            "description": item.description,
+            "version": item.version,
+            "source": item.source,
+            "tools": item.tools,
+            "status": status,
+        })
+
+    await websocket.send(json.dumps({
+        "type": "agent_skills_loaded",
+        "config_name": config_name,
+        "agent_id": agent_id,
+        "skills": payload,
+    }, ensure_ascii=False))
+
+
+async def handle_get_agent_profile(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    if not agent_id:
+        await websocket.send(json.dumps({"type": "error", "message": "get_agent_profile requires agent_id."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    asset_dir = skills_manager.get_agent_asset_dir(config_name, agent_id)
+    agent_config = load_agent_workspace_config(asset_dir / "agent.yaml")
+    profiles = load_agent_profiles()
+    profile = profiles.get(agent_id, {})
+    bootstrap = get_bootstrap_config_for_run(gateway._project_root, config_name)
+    override = bootstrap.agent_override(agent_id)
+    active_tool_groups = override.get("active_tool_groups", agent_config.active_tool_groups or profile.get("active_tool_groups", []))
+    if not isinstance(active_tool_groups, list):
+        active_tool_groups = []
+    disabled_tool_groups = agent_config.disabled_tool_groups
+    if disabled_tool_groups:
+        disabled_set = set(disabled_tool_groups)
+        active_tool_groups = [group_name for group_name in active_tool_groups if group_name not in disabled_set]
+
+    default_skills = profile.get("skills", [])
+    if not isinstance(default_skills, list):
+        default_skills = []
+    resolved_skills = skills_manager.resolve_agent_skill_names(
+        config_name=config_name,
+        agent_id=agent_id,
+        default_skills=default_skills,
+    )
+    prompt_files = agent_config.prompt_files or ["SOUL.md", "PROFILE.md", "AGENTS.md", "POLICY.md", "MEMORY.md"]
+    model_name, model_provider = get_agent_model_info(agent_id)
+
+    await websocket.send(json.dumps({
+        "type": "agent_profile_loaded",
+        "config_name": config_name,
+        "agent_id": agent_id,
+        "profile": {
+            "model_name": model_name,
+            "model_provider": model_provider,
+            "prompt_files": prompt_files,
+            "default_skills": default_skills,
+            "resolved_skills": resolved_skills,
+            "active_tool_groups": active_tool_groups,
+            "disabled_tool_groups": disabled_tool_groups,
+            "enabled_skills": agent_config.enabled_skills,
+            "disabled_skills": agent_config.disabled_skills,
+        },
+    }, ensure_ascii=False))
+
+
+async def handle_get_skill_detail(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    skill_name = str(data.get("skill_name", "")).strip()
+    if not skill_name:
+        await websocket.send(json.dumps({"type": "error", "message": "get_skill_detail requires skill_name."}, ensure_ascii=False))
+        return
+
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    try:
+        if agent_id:
+            config_name = gateway.config.get("config_name", "default")
+            detail = skills_manager.load_agent_skill_document(config_name=config_name, agent_id=agent_id, skill_name=skill_name)
+        else:
+            detail = skills_manager.load_skill_document(skill_name)
+    except FileNotFoundError:
+        await websocket.send(json.dumps({"type": "error", "message": f"Unknown skill: {skill_name}"}, ensure_ascii=False))
+        return
+
+    await websocket.send(json.dumps({
+        "type": "skill_detail_loaded",
+        "agent_id": agent_id,
+        "skill": detail,
+    }, ensure_ascii=False))
+
+
+async def handle_create_agent_local_skill(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    skill_name = str(data.get("skill_name", "")).strip()
+    if not agent_id or not skill_name:
+        await websocket.send(json.dumps({"type": "error", "message": "create_agent_local_skill requires agent_id and skill_name."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    try:
+        skills_manager.create_agent_local_skill(config_name=config_name, agent_id=agent_id, skill_name=skill_name)
+    except (ValueError, FileExistsError) as exc:
+        await websocket.send(json.dumps({"type": "error", "message": str(exc)}, ensure_ascii=False))
+        return
+
+    await gateway.state_sync.on_system_message(f"Created local skill {skill_name} for {agent_id}")
+    await gateway._handle_reload_runtime_assets()
+    await websocket.send(json.dumps({"type": "agent_local_skill_created", "agent_id": agent_id, "skill_name": skill_name}, ensure_ascii=False))
+    await handle_get_agent_skills(gateway, websocket, {"agent_id": agent_id})
+    await handle_get_skill_detail(gateway, websocket, {"agent_id": agent_id, "skill_name": skill_name})
+
+
+async def handle_update_agent_local_skill(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    skill_name = str(data.get("skill_name", "")).strip()
+    content = data.get("content")
+    if not agent_id or not skill_name or not isinstance(content, str):
+        await websocket.send(json.dumps({"type": "error", "message": "update_agent_local_skill requires agent_id, skill_name, and string content."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    try:
+        skills_manager.update_agent_local_skill(config_name=config_name, agent_id=agent_id, skill_name=skill_name, content=content)
+    except (ValueError, FileNotFoundError) as exc:
+        await websocket.send(json.dumps({"type": "error", "message": str(exc)}, ensure_ascii=False))
+        return
+
+    await gateway.state_sync.on_system_message(f"Updated local skill {skill_name} for {agent_id}")
+    await gateway._handle_reload_runtime_assets()
+    await websocket.send(json.dumps({"type": "agent_local_skill_updated", "agent_id": agent_id, "skill_name": skill_name}, ensure_ascii=False))
+    await handle_get_skill_detail(gateway, websocket, {"agent_id": agent_id, "skill_name": skill_name})
+
+
+async def handle_delete_agent_local_skill(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    skill_name = str(data.get("skill_name", "")).strip()
+    if not agent_id or not skill_name:
+        await websocket.send(json.dumps({"type": "error", "message": "delete_agent_local_skill requires agent_id and skill_name."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    try:
+        skills_manager.delete_agent_local_skill(config_name=config_name, agent_id=agent_id, skill_name=skill_name)
+        skills_manager.forget_agent_skill_overrides(config_name=config_name, agent_id=agent_id, skill_names=[skill_name])
+    except (ValueError, FileNotFoundError) as exc:
+        await websocket.send(json.dumps({"type": "error", "message": str(exc)}, ensure_ascii=False))
+        return
+
+    await gateway.state_sync.on_system_message(f"Deleted local skill {skill_name} for {agent_id}")
+    await gateway._handle_reload_runtime_assets()
+    await websocket.send(json.dumps({"type": "agent_local_skill_deleted", "agent_id": agent_id, "skill_name": skill_name}, ensure_ascii=False))
+    await handle_get_agent_skills(gateway, websocket, {"agent_id": agent_id})
+
+
+async def handle_remove_agent_skill(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    skill_name = str(data.get("skill_name", "")).strip()
+    if not agent_id or not skill_name:
+        await websocket.send(json.dumps({"type": "error", "message": "remove_agent_skill requires agent_id and skill_name."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    skill_names = {
+        item.skill_name
+        for item in skills_manager.list_agent_skill_catalog(config_name, agent_id)
+        if item.source != "local"
+    }
+    if skill_name not in skill_names:
+        await websocket.send(json.dumps({"type": "error", "message": f"Unknown shared skill: {skill_name}"}, ensure_ascii=False))
+        return
+
+    skills_manager.update_agent_skill_overrides(config_name=config_name, agent_id=agent_id, disable=[skill_name])
+    await gateway.state_sync.on_system_message(f"Removed shared skill {skill_name} from {agent_id}")
+    await gateway._handle_reload_runtime_assets()
+    await websocket.send(json.dumps({"type": "agent_skill_removed", "agent_id": agent_id, "skill_name": skill_name}, ensure_ascii=False))
+    await handle_get_agent_skills(gateway, websocket, {"agent_id": agent_id})
+
+
+async def handle_update_agent_skill(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    skill_name = str(data.get("skill_name", "")).strip()
+    enabled = data.get("enabled")
+    if not agent_id or not skill_name or not isinstance(enabled, bool):
+        await websocket.send(json.dumps({"type": "error", "message": "update_agent_skill requires agent_id, skill_name, and boolean enabled."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    skill_names = {item.skill_name for item in skills_manager.list_agent_skill_catalog(config_name, agent_id)}
+    if skill_name not in skill_names:
+        await websocket.send(json.dumps({"type": "error", "message": f"Unknown skill: {skill_name}"}, ensure_ascii=False))
+        return
+
+    if enabled:
+        skills_manager.update_agent_skill_overrides(config_name=config_name, agent_id=agent_id, enable=[skill_name])
+        await gateway.state_sync.on_system_message(f"Enabled skill {skill_name} for {agent_id}")
+    else:
+        skills_manager.update_agent_skill_overrides(config_name=config_name, agent_id=agent_id, disable=[skill_name])
+        await gateway.state_sync.on_system_message(f"Disabled skill {skill_name} for {agent_id}")
+
+    await websocket.send(json.dumps({
+        "type": "agent_skill_updated",
+        "agent_id": agent_id,
+        "skill_name": skill_name,
+        "enabled": enabled,
+    }, ensure_ascii=False))
+    await gateway._handle_reload_runtime_assets()
+    await handle_get_agent_skills(gateway, websocket, {"agent_id": agent_id})
+
+
+async def handle_get_agent_workspace_file(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    filename = gateway._normalize_agent_workspace_filename(data.get("filename"))
+    if not agent_id or not filename:
+        await websocket.send(json.dumps({"type": "error", "message": "get_agent_workspace_file requires agent_id and supported filename."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    asset_dir = skills_manager.get_agent_asset_dir(config_name, agent_id)
+    asset_dir.mkdir(parents=True, exist_ok=True)
+    path = asset_dir / filename
+    content = path.read_text(encoding="utf-8") if path.exists() else ""
+    await websocket.send(json.dumps({
+        "type": "agent_workspace_file_loaded",
+        "config_name": config_name,
+        "agent_id": agent_id,
+        "filename": filename,
+        "content": content,
+    }, ensure_ascii=False))
+
+
+async def handle_update_agent_workspace_file(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    agent_id = str(data.get("agent_id", "")).strip()
+    filename = gateway._normalize_agent_workspace_filename(data.get("filename"))
+    content = data.get("content")
+    if not agent_id or not filename or not isinstance(content, str):
+        await websocket.send(json.dumps({"type": "error", "message": "update_agent_workspace_file requires agent_id, supported filename, and string content."}, ensure_ascii=False))
+        return
+
+    config_name = gateway.config.get("config_name", "default")
+    skills_manager = SkillsManager(project_root=gateway._project_root)
+    asset_dir = skills_manager.get_agent_asset_dir(config_name, agent_id)
+    asset_dir.mkdir(parents=True, exist_ok=True)
+    path = asset_dir / filename
+    path.write_text(content, encoding="utf-8")
+    await gateway.state_sync.on_system_message(f"Updated {filename} for {agent_id}")
+    await websocket.send(json.dumps({"type": "agent_workspace_file_updated", "agent_id": agent_id, "filename": filename}, ensure_ascii=False))
+    await gateway._handle_reload_runtime_assets()
+    await handle_get_agent_workspace_file(gateway, websocket, {"agent_id": agent_id, "filename": filename})
--- a/backend/services/gateway_cycle_support.py
+++ b/backend/services/gateway_cycle_support.py
@@ -0,0 +1,372 @@
+# -*- coding: utf-8 -*-
+"""Cycle and monitoring helpers extracted from the main Gateway module."""
+
+from __future__ import annotations
+
+import asyncio
+import logging
+from typing import Any
+
+from backend.data.market_ingest import ingest_symbols
+from backend.tools.data_tools import get_market_cap
+from backend.utils.msg_adapter import FrontendAdapter
+
+logger = logging.getLogger(__name__)
+
+
+def schedule_watchlist_market_store_refresh(gateway: Any, tickers: list[str]) -> None:
+    """Kick off a non-blocking market-store refresh for an updated watchlist."""
+    if not tickers:
+        return
+    if gateway._watchlist_ingest_task and not gateway._watchlist_ingest_task.done():
+        gateway._watchlist_ingest_task.cancel()
+    gateway._watchlist_ingest_task = asyncio.create_task(
+        refresh_market_store_for_watchlist(gateway, tickers),
+    )
+
+
+async def refresh_market_store_for_watchlist(gateway: Any, tickers: list[str]) -> None:
+    """Refresh the long-lived market store after a watchlist update."""
+    try:
+        await gateway.state_sync.on_system_message(
+            f"正在同步自选股市场数据: {', '.join(tickers)}",
+        )
+        results = await asyncio.to_thread(
+            ingest_symbols,
+            tickers,
+            mode="incremental",
+        )
+        summary = ", ".join(
+            f"{item['symbol']} prices={item['prices']} news={item['news']}"
+            for item in results
+        )
+        await gateway.state_sync.on_system_message(
+            f"自选股市场数据已同步: {summary}",
+        )
+    except asyncio.CancelledError:
+        raise
+    except Exception as exc:
+        logger.warning("Watchlist market store refresh failed: %s", exc)
+        await gateway.state_sync.on_system_message(
+            f"自选股市场数据同步失败: {exc}",
+        )
+
+
+async def market_status_monitor(gateway: Any) -> None:
+    """Periodically check and broadcast market status changes."""
+    while True:
+        try:
+            await gateway.market_service.check_and_broadcast_market_status()
+
+            status = gateway.market_service.get_market_status()
+            if status["status"] == "open" and not gateway.storage.is_live_session_active:
+                gateway.storage.start_live_session()
+                summary = gateway.storage.load_file("summary") or {}
+                gateway._session_start_portfolio_value = summary.get(
+                    "totalAssetValue",
+                    gateway.storage.initial_cash,
+                )
+                logger.info(
+                    "Session start portfolio: $%s",
+                    f"{gateway._session_start_portfolio_value:,.2f}",
+                )
+            elif status["status"] != "open" and gateway.storage.is_live_session_active:
+                gateway.storage.end_live_session()
+                gateway._session_start_portfolio_value = None
+
+            if gateway.storage.is_live_session_active:
+                await update_and_broadcast_live_returns(gateway)
+
+            await asyncio.sleep(60)
+        except asyncio.CancelledError:
+            break
+        except Exception as exc:
+            logger.error("Market status monitor error: %s", exc)
+            await asyncio.sleep(60)
+
+
+async def update_and_broadcast_live_returns(gateway: Any) -> None:
+    """Calculate and broadcast live returns for current session."""
+    if not gateway.storage.is_live_session_active:
+        return
+
+    prices = gateway.market_service.get_all_prices()
+    if not prices or not any(p > 0 for p in prices.values()):
+        return
+
+    state = gateway.storage.load_internal_state()
+    equity_history = state.get("equity_history", [])
+    baseline_history = state.get("baseline_history", [])
+    baseline_vw_history = state.get("baseline_vw_history", [])
+    momentum_history = state.get("momentum_history", [])
+
+    current_equity = equity_history[-1]["v"] if equity_history else None
+    current_baseline = baseline_history[-1]["v"] if baseline_history else None
+    current_baseline_vw = baseline_vw_history[-1]["v"] if baseline_vw_history else None
+    current_momentum = momentum_history[-1]["v"] if momentum_history else None
+
+    point = gateway.storage.update_live_returns(
+        current_equity=current_equity,
+        current_baseline=current_baseline,
+        current_baseline_vw=current_baseline_vw,
+        current_momentum=current_momentum,
+    )
+    if point:
+        live_returns = gateway.storage.get_live_returns()
+        await gateway.broadcast(
+            {
+                "type": "team_summary",
+                "equity_return": live_returns["equity_return"],
+                "baseline_return": live_returns["baseline_return"],
+                "baseline_vw_return": live_returns["baseline_vw_return"],
+                "momentum_return": live_returns["momentum_return"],
+            },
+        )
+
+
+async def on_strategy_trigger(gateway: Any, date: str) -> None:
+    """Handle trading cycle trigger."""
+    if gateway._cycle_lock.locked():
+        logger.warning("Trading cycle already running, skipping trigger for %s", date)
+        await gateway.state_sync.on_system_message(f"已有交易周期在运行，跳过本次触发: {date}")
+        return
+
+    async with gateway._cycle_lock:
+        logger.info("Strategy triggered for %s", date)
+        tickers = gateway.config.get("tickers", [])
+        if gateway.is_backtest:
+            await run_backtest_cycle(gateway, date, tickers)
+        else:
+            await run_live_cycle(gateway, date, tickers)
+
+
+async def on_heartbeat_trigger(gateway: Any, date: str) -> None:
+    """Run lightweight heartbeat check for all analysts."""
+    logger.info("[Heartbeat] Running heartbeat check for %s", date)
+    analysts = gateway.pipeline._all_analysts()
+
+    for analyst in analysts:
+        try:
+            ws_id = getattr(analyst, "workspace_id", None)
+            if ws_id:
+                from backend.agents.workspace_manager import get_workspace_dir
+                from pathlib import Path
+                from agentscope.message import Msg
+
+                ws_dir = get_workspace_dir(ws_id)
+                if ws_dir:
+                    hb_path = Path(ws_dir) / "HEARTBEAT.md"
+                    if hb_path.exists():
+                        content = hb_path.read_text(encoding="utf-8").strip()
+                        if content:
+                            hb_task = f"# 定期主动检查\n\n{content}\n\n请执行上述检查并报告结果。"
+                            logger.info("[Heartbeat] Running heartbeat for %s", analyst.name)
+                            msg = Msg(role="user", content=hb_task, name="system")
+                            await analyst.reply([msg])
+                            logger.info("[Heartbeat] %s heartbeat complete", analyst.name)
+                            continue
+            logger.debug("[Heartbeat] No HEARTBEAT.md for %s, skipping", analyst.name)
+        except Exception as exc:
+            logger.error("[Heartbeat] %s failed: %s", analyst.name, exc, exc_info=True)
+
+
+async def run_backtest_cycle(gateway: Any, date: str, tickers: list[str]) -> None:
+    gateway.market_service.set_backtest_date(date)
+    await gateway.market_service.emit_market_open()
+
+    await gateway.state_sync.on_cycle_start(date)
+    gateway._dashboard.update(date=date, status="Analyzing...")
+
+    prices = gateway.market_service.get_open_prices()
+    close_prices = gateway.market_service.get_close_prices()
+    market_caps = await get_market_caps(gateway, tickers, date)
+
+    result = await gateway.pipeline.run_cycle(
+        tickers=tickers,
+        date=date,
+        prices=prices,
+        close_prices=close_prices,
+        market_caps=market_caps,
+    )
+
+    await gateway.market_service.emit_market_close()
+    settlement_result = result.get("settlement_result")
+    save_cycle_results(gateway, result, date, close_prices, settlement_result)
+    await broadcast_portfolio_updates(gateway, result, close_prices)
+    await finalize_cycle(gateway, date)
+
+
+async def run_live_cycle(gateway: Any, date: str, tickers: list[str]) -> None:
+    trading_date = gateway.market_service.get_live_trading_date()
+    logger.info("Live cycle: triggered=%s, trading_date=%s", date, trading_date)
+
+    await gateway.state_sync.on_cycle_start(trading_date)
+    gateway._dashboard.update(date=trading_date, status="Analyzing...")
+
+    market_caps = await get_market_caps(gateway, tickers, trading_date)
+    schedule_mode = gateway.config.get("schedule_mode", "daily")
+    market_status = gateway.market_service.get_market_status()
+    current_prices = gateway.market_service.get_all_prices()
+
+    if schedule_mode == "intraday":
+        execute_decisions = market_status.get("status") == "open"
+        if execute_decisions:
+            await gateway.state_sync.on_system_message("定时任务触发：当前处于交易时段，本轮将执行交易决策")
+        else:
+            await gateway.state_sync.on_system_message("定时任务触发：当前非交易时段，本轮仅更新数据与分析，不执行交易")
+
+        result = await gateway.pipeline.run_cycle(
+            tickers=tickers,
+            date=trading_date,
+            prices=current_prices,
+            market_caps=market_caps,
+            execute_decisions=execute_decisions,
+        )
+        close_prices = current_prices
+    else:
+        result = await gateway.pipeline.run_cycle(
+            tickers=tickers,
+            date=trading_date,
+            market_caps=market_caps,
+            get_open_prices_fn=gateway.market_service.wait_for_open_prices,
+            get_close_prices_fn=gateway.market_service.wait_for_close_prices,
+        )
+        close_prices = gateway.market_service.get_all_prices()
+
+    settlement_result = result.get("settlement_result")
+    save_cycle_results(gateway, result, trading_date, close_prices, settlement_result)
+    await broadcast_portfolio_updates(gateway, result, close_prices)
+    await finalize_cycle(gateway, trading_date)
+
+
+async def finalize_cycle(gateway: Any, date: str) -> None:
+    summary = gateway.storage.load_file("summary") or {}
+    if gateway.storage.is_live_session_active:
+        summary.update(gateway.storage.get_live_returns())
+
+    await gateway.state_sync.on_cycle_end(date, portfolio_summary=summary)
+    holdings = gateway.storage.load_file("holdings") or []
+    trades = gateway.storage.load_file("trades") or []
+    leaderboard = gateway.storage.load_file("leaderboard") or []
+    if leaderboard:
+        await gateway.state_sync.on_leaderboard_update(leaderboard)
+    gateway._dashboard.update(date=date, status="Running", portfolio=summary, holdings=holdings, trades=trades)
+
+
+async def get_market_caps(gateway: Any, tickers: list[str], date: str) -> dict[str, float]:
+    market_caps: dict[str, float] = {}
+    for ticker in tickers:
+        try:
+            market_cap = None
+            response = await gateway._call_trading_service(
+                f"get_market_cap for {ticker}",
+                lambda client, symbol=ticker: client.get_market_cap(ticker=symbol, end_date=date),
+            )
+            if response is not None:
+                market_cap = response.get("market_cap")
+            if market_cap is None:
+                market_cap = get_market_cap(ticker=ticker, end_date=date)
+            market_caps[ticker] = market_cap if market_cap else 1e9
+        except Exception as exc:
+            logger.warning("Failed to get market cap for %s, using default 1e9: %s", ticker, exc)
+            market_caps[ticker] = 1e9
+    return market_caps
+
+
+async def broadcast_portfolio_updates(gateway: Any, result: dict[str, Any], prices: dict[str, float]) -> None:
+    portfolio = result.get("portfolio", {})
+    if portfolio:
+        holdings = FrontendAdapter.build_holdings(portfolio, prices)
+        if holdings:
+            await gateway.state_sync.on_holdings_update(holdings)
+        stats = FrontendAdapter.build_stats(portfolio, prices)
+        if stats:
+            await gateway.state_sync.on_stats_update(stats)
+
+    executed_trades = result.get("executed_trades", [])
+    if executed_trades:
+        await gateway.state_sync.on_trades_executed(executed_trades)
+
+
+def save_cycle_results(
+    gateway: Any,
+    result: dict[str, Any],
+    date: str,
+    prices: dict[str, float],
+    settlement_result: dict[str, Any] | None = None,
+) -> None:
+    portfolio = result.get("portfolio", {})
+    executed_trades = result.get("executed_trades", [])
+    baseline_values = settlement_result.get("baseline_values") if settlement_result else None
+    if portfolio:
+        gateway.storage.update_dashboard_after_cycle(
+            portfolio=portfolio,
+            prices=prices,
+            date=date,
+            executed_trades=executed_trades,
+            baseline_values=baseline_values,
+        )
+
+
+async def run_backtest_dates(gateway: Any, dates: list[str]) -> None:
+    gateway.state_sync.set_backtest_dates(dates)
+    gateway._dashboard.update(days_total=len(dates), days_completed=0)
+    await gateway.state_sync.on_system_message(f"Starting backtest - {len(dates)} trading days")
+    try:
+        for i, date in enumerate(dates):
+            gateway._dashboard.update(days_completed=i)
+            await gateway.on_strategy_trigger(date=date)
+            await asyncio.sleep(0.1)
+        await gateway.state_sync.on_system_message(f"Backtest complete - {len(dates)} days")
+        summary = gateway.storage.load_file("summary") or {}
+        gateway._dashboard.update(status="Complete", portfolio=summary, days_completed=len(dates))
+        gateway._dashboard.stop()
+        gateway._dashboard.print_final_summary()
+    except Exception as exc:
+        error_msg = f"Backtest failed: {type(exc).__name__}: {str(exc)}"
+        logger.error(error_msg, exc_info=True)
+        asyncio.create_task(gateway.state_sync.on_system_message(error_msg))
+        gateway._dashboard.update(status=f"Failed: {str(exc)}")
+        gateway._dashboard.stop()
+        raise
+    finally:
+        gateway._backtest_task = None
+
+
+def handle_backtest_exception(gateway: Any, task: asyncio.Task) -> None:
+    try:
+        task.result()
+    except asyncio.CancelledError:
+        logger.info("Backtest task was cancelled")
+    except Exception as exc:
+        logger.error("Backtest task failed with exception:%s:%s", type(exc).__name__, exc, exc_info=True)
+
+
+def handle_manual_cycle_exception(gateway: Any, task: asyncio.Task) -> None:
+    gateway._manual_cycle_task = None
+    try:
+        task.result()
+    except asyncio.CancelledError:
+        logger.info("Manual cycle task was cancelled")
+    except Exception as exc:
+        logger.error("Manual cycle task failed with exception:%s:%s", type(exc).__name__, exc, exc_info=True)
+
+
+def set_backtest_dates(gateway: Any, dates: list[str]) -> None:
+    gateway.state_sync.set_backtest_dates(dates)
+    if dates:
+        gateway._backtest_start_date = dates[0]
+        gateway._backtest_end_date = dates[-1]
+        gateway._dashboard.days_total = len(dates)
+
+
+def stop_gateway(gateway: Any) -> None:
+    gateway.state_sync.save_state()
+    gateway.market_service.stop()
+    if gateway._backtest_task:
+        gateway._backtest_task.cancel()
+    if gateway._market_status_task:
+        gateway._market_status_task.cancel()
+    if gateway._watchlist_ingest_task:
+        gateway._watchlist_ingest_task.cancel()
+    gateway._dashboard.stop()
--- a/backend/services/gateway_runtime_support.py
+++ b/backend/services/gateway_runtime_support.py
@@ -0,0 +1,174 @@
+# -*- coding: utf-8 -*-
+"""Runtime/state support helpers extracted from the main Gateway module."""
+
+from __future__ import annotations
+
+from typing import Any
+
+from backend.data.provider_utils import normalize_symbol
+
+
+def normalize_watchlist(raw_tickers: Any) -> list[str]:
+    """Parse watchlist payloads from websocket messages."""
+    if raw_tickers is None:
+        return []
+
+    if isinstance(raw_tickers, str):
+        candidates = raw_tickers.split(",")
+    elif isinstance(raw_tickers, list):
+        candidates = raw_tickers
+    else:
+        candidates = [raw_tickers]
+
+    tickers: list[str] = []
+    for candidate in candidates:
+        symbol = normalize_symbol(str(candidate).strip().strip("\"'"))
+        if symbol and symbol not in tickers:
+            tickers.append(symbol)
+    return tickers
+
+
+def normalize_agent_workspace_filename(
+    raw_name: Any,
+    *,
+    allowlist: set[str],
+) -> str | None:
+    """Restrict editable workspace files to a safe allowlist."""
+    filename = str(raw_name or "").strip()
+    if filename in allowlist:
+        return filename
+    return None
+
+
+def apply_runtime_config(gateway: Any, runtime_config: dict[str, Any]) -> dict[str, Any]:
+    """Apply runtime config to gateway-owned services and state."""
+    warnings: list[str] = []
+
+    ticker_changes = gateway.market_service.update_tickers(
+        runtime_config.get("tickers", []),
+    )
+    gateway.config["tickers"] = ticker_changes["active"]
+
+    gateway.pipeline.max_comm_cycles = int(runtime_config["max_comm_cycles"])
+    gateway.config["max_comm_cycles"] = gateway.pipeline.max_comm_cycles
+    gateway.config["schedule_mode"] = runtime_config.get(
+        "schedule_mode",
+        gateway.config.get("schedule_mode", "daily"),
+    )
+    gateway.config["interval_minutes"] = int(
+        runtime_config.get(
+            "interval_minutes",
+            gateway.config.get("interval_minutes", 60),
+        ),
+    )
+    gateway.config["trigger_time"] = runtime_config.get(
+        "trigger_time",
+        gateway.config.get("trigger_time", "09:30"),
+    )
+
+    if gateway.scheduler:
+        gateway.scheduler.reconfigure(
+            mode=gateway.config["schedule_mode"],
+            trigger_time=gateway.config["trigger_time"],
+            interval_minutes=gateway.config["interval_minutes"],
+        )
+
+    pm_apply_result = gateway.pipeline.pm.apply_runtime_portfolio_config(
+        margin_requirement=runtime_config["margin_requirement"],
+    )
+    gateway.config["margin_requirement"] = gateway.pipeline.pm.portfolio.get(
+        "margin_requirement",
+        runtime_config["margin_requirement"],
+    )
+
+    requested_initial_cash = float(runtime_config["initial_cash"])
+    current_initial_cash = float(gateway.storage.initial_cash)
+    initial_cash_applied = requested_initial_cash == current_initial_cash
+    if not initial_cash_applied:
+        if (
+            gateway.storage.can_apply_initial_cash()
+            and gateway.pipeline.pm.can_apply_initial_cash()
+        ):
+            initial_cash_applied = gateway.storage.apply_initial_cash(
+                requested_initial_cash,
+            )
+            if initial_cash_applied:
+                gateway.pipeline.pm.apply_runtime_portfolio_config(
+                    initial_cash=requested_initial_cash,
+                )
+                gateway.config["initial_cash"] = gateway.storage.initial_cash
+        else:
+            warnings.append(
+                "initial_cash changed in BOOTSTRAP.md but was not applied "
+                "because the run already has positions, margin usage, or trades.",
+            )
+
+    requested_enable_memory = bool(runtime_config["enable_memory"])
+    current_enable_memory = bool(gateway.config.get("enable_memory", False))
+    if requested_enable_memory != current_enable_memory:
+        warnings.append(
+            "enable_memory changed in BOOTSTRAP.md but still requires a restart "
+            "because long-term memory contexts are created at startup.",
+        )
+
+    sync_runtime_state(gateway)
+
+    return {
+        "runtime_config_requested": runtime_config,
+        "runtime_config_applied": {
+            "tickers": list(gateway.config.get("tickers", [])),
+            "schedule_mode": gateway.config.get("schedule_mode", "daily"),
+            "interval_minutes": gateway.config.get("interval_minutes", 60),
+            "trigger_time": gateway.config.get("trigger_time", "09:30"),
+            "initial_cash": gateway.storage.initial_cash,
+            "margin_requirement": gateway.config["margin_requirement"],
+            "max_comm_cycles": gateway.config["max_comm_cycles"],
+            "enable_memory": gateway.config.get("enable_memory", False),
+        },
+        "runtime_config_status": {
+            "tickers": True,
+            "schedule_mode": True,
+            "interval_minutes": True,
+            "trigger_time": True,
+            "initial_cash": initial_cash_applied,
+            "margin_requirement": pm_apply_result["margin_requirement"],
+            "max_comm_cycles": True,
+            "enable_memory": requested_enable_memory == current_enable_memory,
+        },
+        "ticker_changes": ticker_changes,
+        "runtime_config_warnings": warnings,
+    }
+
+
+def sync_runtime_state(gateway: Any) -> None:
+    """Refresh persisted state and dashboard after runtime config changes."""
+    gateway.state_sync.update_state("tickers", gateway.config.get("tickers", []))
+    gateway.state_sync.update_state(
+        "runtime_config",
+        {
+            "tickers": gateway.config.get("tickers", []),
+            "schedule_mode": gateway.config.get("schedule_mode", "daily"),
+            "interval_minutes": gateway.config.get("interval_minutes", 60),
+            "trigger_time": gateway.config.get("trigger_time", "09:30"),
+            "initial_cash": gateway.storage.initial_cash,
+            "margin_requirement": gateway.config.get("margin_requirement"),
+            "max_comm_cycles": gateway.config.get("max_comm_cycles"),
+            "enable_memory": gateway.config.get("enable_memory", False),
+        },
+    )
+
+    gateway.storage.update_server_state_from_dashboard(gateway.state_sync.state)
+    gateway.state_sync.save_state()
+
+    gateway._dashboard.tickers = list(gateway.config.get("tickers", []))
+    gateway._dashboard.initial_cash = gateway.storage.initial_cash
+    gateway._dashboard.enable_memory = bool(gateway.config.get("enable_memory", False))
+
+    summary = gateway.storage.load_file("summary") or {}
+    holdings = gateway.storage.load_file("holdings") or []
+    trades = gateway.storage.load_file("trades") or []
+    gateway._dashboard.update(
+        portfolio=summary,
+        holdings=holdings,
+        trades=trades,
+    )
--- a/backend/services/gateway_stock_handlers.py
+++ b/backend/services/gateway_stock_handlers.py
@@ -0,0 +1,707 @@
+# -*- coding: utf-8 -*-
+"""Stock-related Gateway handlers extracted from the main Gateway module."""
+
+from __future__ import annotations
+
+import asyncio
+import json
+import logging
+from datetime import datetime, timedelta
+from typing import Any
+
+from backend.data.provider_utils import normalize_symbol
+from backend.domains import news as news_domain
+from backend.enrich.news_enricher import enrich_news_for_symbol
+from backend.enrich.llm_enricher import llm_enrichment_enabled
+from backend.tools.data_tools import get_insider_trades, get_prices, prices_to_df
+from shared.client import NewsServiceClient, TradingServiceClient
+
+logger = logging.getLogger(__name__)
+
+
+async def handle_get_stock_history(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    if not ticker:
+        await websocket.send(json.dumps({
+            "type": "stock_history_loaded",
+            "ticker": "",
+            "prices": [],
+            "source": None,
+            "error": "invalid ticker",
+        }, ensure_ascii=False))
+        return
+
+    lookback_days = data.get("lookback_days", 90)
+    try:
+        lookback_days = max(7, min(int(lookback_days), 365))
+    except (TypeError, ValueError):
+        lookback_days = 90
+
+    end_date = gateway.state_sync.state.get("current_date") or datetime.now().strftime("%Y-%m-%d")
+    try:
+        end_dt = datetime.strptime(end_date, "%Y-%m-%d")
+    except ValueError:
+        end_dt = datetime.now()
+        end_date = end_dt.strftime("%Y-%m-%d")
+    start_date = (end_dt - timedelta(days=lookback_days)).strftime("%Y-%m-%d")
+
+    prices = []
+    source = "polygon"
+    response = await gateway._call_trading_service(
+        "get_prices for history",
+        lambda client: client.get_prices(ticker=ticker, start_date=start_date, end_date=end_date),
+    )
+    if response is not None:
+        prices = response.prices
+        source = "trading_service"
+
+    if not prices:
+        prices = await asyncio.to_thread(gateway.storage.market_store.get_ohlc, ticker, start_date, end_date)
+        if not prices:
+            prices = await asyncio.to_thread(
+                get_prices,
+                ticker=ticker,
+                start_date=start_date,
+                end_date=end_date,
+            )
+            usage_snapshot = gateway._provider_router.get_usage_snapshot()
+            source = usage_snapshot.get("last_success", {}).get("prices")
+            if prices:
+                await asyncio.to_thread(
+                    gateway.storage.market_store.upsert_ohlc,
+                    ticker,
+                    [price.model_dump() for price in prices],
+                    source=source or "provider",
+                )
+
+    await websocket.send(json.dumps({
+        "type": "stock_history_loaded",
+        "ticker": ticker,
+        "prices": [price if isinstance(price, dict) else price.model_dump() for price in prices][-120:],
+        "source": source,
+        "start_date": start_date,
+        "end_date": end_date,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_explain_events(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    snapshot = gateway.storage.runtime_db.get_stock_explain_snapshot(ticker)
+    await websocket.send(json.dumps({
+        "type": "stock_explain_events_loaded",
+        "ticker": ticker,
+        "events": snapshot.get("events", []),
+        "signals": snapshot.get("signals", []),
+        "trades": snapshot.get("trades", []),
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_news(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    if not ticker:
+        await websocket.send(json.dumps({
+            "type": "stock_news_loaded",
+            "ticker": "",
+            "news": [],
+            "source": None,
+            "error": "invalid ticker",
+        }, ensure_ascii=False))
+        return
+
+    lookback_days = data.get("lookback_days", 30)
+    limit = data.get("limit", 12)
+    try:
+        lookback_days = max(7, min(int(lookback_days), 180))
+    except (TypeError, ValueError):
+        lookback_days = 30
+    try:
+        limit = max(1, min(int(limit), 30))
+    except (TypeError, ValueError):
+        limit = 12
+
+    end_date = gateway.state_sync.state.get("current_date") or datetime.now().strftime("%Y-%m-%d")
+    try:
+        end_dt = datetime.strptime(end_date, "%Y-%m-%d")
+    except ValueError:
+        end_dt = datetime.now()
+        end_date = end_dt.strftime("%Y-%m-%d")
+    start_date = (end_dt - timedelta(days=lookback_days)).strftime("%Y-%m-%d")
+
+    news_rows = []
+    source = "polygon"
+    response = await gateway._call_news_service(
+        "get_enriched_news",
+        lambda client: client.get_enriched_news(
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=limit,
+        ),
+    )
+    if response is not None:
+        news_rows = response.get("news") or []
+        source = "news_service"
+
+    if not news_rows:
+        payload = await asyncio.to_thread(
+            news_domain.get_enriched_news,
+            gateway.storage.market_store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=max(limit, 50),
+        )
+        news_rows = (payload.get("news") or [])[-limit:]
+        source = "market_store"
+
+    await websocket.send(json.dumps({
+        "type": "stock_news_loaded",
+        "ticker": ticker,
+        "news": news_rows[-limit:],
+        "source": source,
+        "start_date": start_date,
+        "end_date": end_date,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_news_for_date(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    trade_date = str(data.get("date") or "").strip()
+    if not ticker or not trade_date:
+        await websocket.send(json.dumps({
+            "type": "stock_news_for_date_loaded",
+            "ticker": ticker,
+            "date": trade_date,
+            "news": [],
+            "error": "ticker and date are required",
+        }, ensure_ascii=False))
+        return
+
+    limit = data.get("limit", 20)
+    try:
+        limit = max(1, min(int(limit), 50))
+    except (TypeError, ValueError):
+        limit = 20
+
+    source = "market_store"
+    news_rows = []
+    response = await gateway._call_news_service(
+        "get_news_for_date",
+        lambda client: client.get_news_for_date(ticker=ticker, date=trade_date, limit=limit),
+    )
+    if response is not None:
+        news_rows = response.get("news") or []
+        source = "news_service"
+
+    if not news_rows:
+        payload = await asyncio.to_thread(
+            news_domain.get_news_for_date,
+            gateway.storage.market_store,
+            ticker=ticker,
+            date=trade_date,
+            limit=limit,
+        )
+        news_rows = payload.get("news") or []
+        source = "market_store"
+
+    await websocket.send(json.dumps({
+        "type": "stock_news_for_date_loaded",
+        "ticker": ticker,
+        "date": trade_date,
+        "news": news_rows,
+        "source": source,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_news_timeline(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    if not ticker:
+        await websocket.send(json.dumps({
+            "type": "stock_news_timeline_loaded",
+            "ticker": "",
+            "timeline": [],
+            "error": "invalid ticker",
+        }, ensure_ascii=False))
+        return
+
+    lookback_days = data.get("lookback_days", 90)
+    try:
+        lookback_days = max(7, min(int(lookback_days), 365))
+    except (TypeError, ValueError):
+        lookback_days = 90
+
+    end_date = gateway.state_sync.state.get("current_date") or datetime.now().strftime("%Y-%m-%d")
+    try:
+        end_dt = datetime.strptime(end_date, "%Y-%m-%d")
+    except ValueError:
+        end_dt = datetime.now()
+        end_date = end_dt.strftime("%Y-%m-%d")
+    start_date = (end_dt - timedelta(days=lookback_days)).strftime("%Y-%m-%d")
+
+    timeline = []
+    response = await gateway._call_news_service(
+        "get_news_timeline",
+        lambda client: client.get_news_timeline(ticker=ticker, start_date=start_date, end_date=end_date),
+    )
+    if response is not None:
+        timeline = response.get("timeline") or []
+
+    if not timeline:
+        payload = await asyncio.to_thread(
+            news_domain.get_news_timeline,
+            gateway.storage.market_store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+        )
+        timeline = payload.get("timeline") or []
+
+    await websocket.send(json.dumps({
+        "type": "stock_news_timeline_loaded",
+        "ticker": ticker,
+        "timeline": timeline,
+        "start_date": start_date,
+        "end_date": end_date,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_news_categories(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    if not ticker:
+        await websocket.send(json.dumps({
+            "type": "stock_news_categories_loaded",
+            "ticker": "",
+            "categories": {},
+            "error": "invalid ticker",
+        }, ensure_ascii=False))
+        return
+
+    lookback_days = data.get("lookback_days", 90)
+    try:
+        lookback_days = max(7, min(int(lookback_days), 365))
+    except (TypeError, ValueError):
+        lookback_days = 90
+
+    end_date = gateway.state_sync.state.get("current_date") or datetime.now().strftime("%Y-%m-%d")
+    try:
+        end_dt = datetime.strptime(end_date, "%Y-%m-%d")
+    except ValueError:
+        end_dt = datetime.now()
+        end_date = end_dt.strftime("%Y-%m-%d")
+    start_date = (end_dt - timedelta(days=lookback_days)).strftime("%Y-%m-%d")
+
+    categories = {}
+    response = await gateway._call_news_service(
+        "get_categories",
+        lambda client: client.get_categories(
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=200,
+        ),
+    )
+    if response is not None:
+        categories = response.get("categories") or {}
+
+    if not categories:
+        payload = await asyncio.to_thread(
+            news_domain.get_news_categories,
+            gateway.storage.market_store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            limit=200,
+        )
+        categories = payload.get("categories") or {}
+
+    await websocket.send(json.dumps({
+        "type": "stock_news_categories_loaded",
+        "ticker": ticker,
+        "categories": categories,
+        "start_date": start_date,
+        "end_date": end_date,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_range_explain(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    start_date = str(data.get("start_date") or "").strip()
+    end_date = str(data.get("end_date") or "").strip()
+    if not ticker or not start_date or not end_date:
+        await websocket.send(json.dumps({
+            "type": "stock_range_explain_loaded",
+            "ticker": ticker,
+            "result": {"error": "ticker, start_date, end_date are required"},
+        }, ensure_ascii=False))
+        return
+
+    article_ids = data.get("article_ids")
+    result = None
+    response = await gateway._call_news_service(
+        "get_range_explain",
+        lambda client: client.get_range_explain(
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            article_ids=article_ids if isinstance(article_ids, list) else None,
+            limit=100,
+        ),
+    )
+    if response is not None:
+        result = response.get("result")
+
+    if result is None:
+        payload = await asyncio.to_thread(
+            news_domain.get_range_explain_payload,
+            gateway.storage.market_store,
+            ticker=ticker,
+            start_date=start_date,
+            end_date=end_date,
+            article_ids=article_ids if isinstance(article_ids, list) else None,
+            limit=100,
+        )
+        result = payload.get("result")
+
+    await websocket.send(json.dumps({
+        "type": "stock_range_explain_loaded",
+        "ticker": ticker,
+        "result": result,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_insider_trades(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    if not ticker:
+        await websocket.send(json.dumps({
+            "type": "stock_insider_trades_loaded",
+            "ticker": "",
+            "trades": [],
+            "error": "invalid ticker",
+        }, ensure_ascii=False))
+        return
+
+    end_date = str(data.get("end_date") or gateway.state_sync.state.get("current_date") or datetime.now().strftime("%Y-%m-%d")).strip()[:10]
+    start_date = str(data.get("start_date") or "").strip()[:10]
+    limit = int(data.get("limit", 50))
+
+    trades = []
+    response = await gateway._call_trading_service(
+        "get_insider_trades",
+        lambda client: client.get_insider_trades(
+            ticker=ticker,
+            end_date=end_date,
+            start_date=start_date if start_date else None,
+            limit=limit,
+        ),
+    )
+    if response is not None:
+        trades = response.insider_trades
+
+    if not trades:
+        trades = await asyncio.to_thread(
+            get_insider_trades,
+            ticker=ticker,
+            end_date=end_date,
+            start_date=start_date if start_date else None,
+            limit=limit,
+        )
+
+    sorted_trades = sorted(trades, key=lambda t: t.transaction_date or "", reverse=True)
+    formatted_trades = [{
+        "ticker": t.ticker,
+        "name": t.name,
+        "title": t.title,
+        "is_board_director": t.is_board_director,
+        "transaction_date": t.transaction_date,
+        "transaction_shares": t.transaction_shares,
+        "transaction_price_per_share": t.transaction_price_per_share,
+        "transaction_value": t.transaction_value,
+        "shares_owned_before_transaction": t.shares_owned_before_transaction,
+        "shares_owned_after_transaction": t.shares_owned_after_transaction,
+        "security_title": t.security_title,
+        "filing_date": t.filing_date,
+        "holding_change": (
+            (t.shares_owned_after_transaction or 0) - (t.shares_owned_before_transaction or 0)
+            if t.shares_owned_after_transaction and t.shares_owned_before_transaction else None
+        ),
+        "is_buy": ((t.transaction_shares or 0) > 0) if t.transaction_shares is not None else None,
+    } for t in sorted_trades]
+
+    await websocket.send(json.dumps({
+        "type": "stock_insider_trades_loaded",
+        "ticker": ticker,
+        "start_date": start_date or None,
+        "end_date": end_date,
+        "trades": formatted_trades,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_story(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    if not ticker:
+        await websocket.send(json.dumps({
+            "type": "stock_story_loaded",
+            "ticker": "",
+            "story": "",
+            "error": "invalid ticker",
+        }, ensure_ascii=False))
+        return
+
+    as_of_date = str(data.get("as_of_date") or gateway.state_sync.state.get("current_date") or datetime.now().strftime("%Y-%m-%d")).strip()[:10]
+    result = await gateway._call_news_service(
+        "get_story",
+        lambda client: client.get_story(ticker=ticker, as_of_date=as_of_date),
+    )
+    if result is None:
+        result = await asyncio.to_thread(
+            news_domain.get_story_payload,
+            gateway.storage.market_store,
+            ticker=ticker,
+            as_of_date=as_of_date,
+        )
+
+    await websocket.send(json.dumps({
+        "type": "stock_story_loaded",
+        "ticker": ticker,
+        "as_of_date": as_of_date,
+        "story": result.get("story") or "",
+        "source": result.get("source") or "local",
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_similar_days(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    target_date = str(data.get("date") or "").strip()[:10]
+    if not ticker or not target_date:
+        await websocket.send(json.dumps({
+            "type": "stock_similar_days_loaded",
+            "ticker": ticker,
+            "date": target_date,
+            "items": [],
+            "error": "ticker and date are required",
+        }, ensure_ascii=False))
+        return
+
+    top_k = data.get("top_k", 8)
+    try:
+        top_k = max(1, min(int(top_k), 20))
+    except (TypeError, ValueError):
+        top_k = 8
+
+    result = await gateway._call_news_service(
+        "get_similar_days",
+        lambda client: client.get_similar_days(ticker=ticker, date=target_date, n_similar=top_k),
+    )
+    if result is None:
+        result = await asyncio.to_thread(
+            news_domain.get_similar_days_payload,
+            gateway.storage.market_store,
+            ticker=ticker,
+            date=target_date,
+            n_similar=top_k,
+        )
+
+    await websocket.send(json.dumps({
+        "type": "stock_similar_days_loaded",
+        "ticker": ticker,
+        "date": target_date,
+        **result,
+    }, ensure_ascii=False, default=str))
+
+
+async def handle_get_stock_technical_indicators(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    if not ticker:
+        await websocket.send(json.dumps({
+            "type": "stock_technical_indicators_loaded",
+            "ticker": ticker,
+            "indicators": None,
+            "error": "ticker is required",
+        }, ensure_ascii=False))
+        return
+
+    try:
+        end_date = datetime.now()
+        start_date = end_date - timedelta(days=250)
+
+        prices = None
+        response = await gateway._call_trading_service(
+            "get_prices",
+            lambda client: client.get_prices(
+                ticker=ticker,
+                start_date=start_date.strftime("%Y-%m-%d"),
+                end_date=end_date.strftime("%Y-%m-%d"),
+            ),
+        )
+        if response is not None:
+            prices = response.prices
+
+        if prices is None:
+            prices = get_prices(
+                ticker=ticker,
+                start_date=start_date.strftime("%Y-%m-%d"),
+                end_date=end_date.strftime("%Y-%m-%d"),
+            )
+
+        if not prices or len(prices) < 20:
+            await websocket.send(json.dumps({
+                "type": "stock_technical_indicators_loaded",
+                "ticker": ticker,
+                "indicators": None,
+                "error": "Insufficient price data",
+            }, ensure_ascii=False))
+            return
+
+        df = prices_to_df(prices)
+        signal = gateway._technical_analyzer.analyze(ticker, df)
+
+        import pandas as pd
+        df_sorted = df.sort_values("time").reset_index(drop=True)
+        df_sorted["returns"] = df_sorted["close"].pct_change()
+        vol_10 = float(df_sorted["returns"].tail(10).std() * (252**0.5) * 100) if len(df_sorted) >= 10 else None
+        vol_20 = float(df_sorted["returns"].tail(20).std() * (252**0.5) * 100) if len(df_sorted) >= 20 else None
+        vol_60 = float(df_sorted["returns"].tail(60).std() * (252**0.5) * 100) if len(df_sorted) >= 60 else None
+        ma_distance = {}
+        for ma_key in ["ma5", "ma10", "ma20", "ma50", "ma200"]:
+            ma_value = getattr(signal, ma_key, None)
+            ma_distance[ma_key] = ((signal.current_price - ma_value) / ma_value) * 100 if ma_value and ma_value > 0 else None
+
+        indicators = {
+            "ticker": ticker,
+            "current_price": signal.current_price,
+            "ma": {
+                "ma5": signal.ma5,
+                "ma10": signal.ma10,
+                "ma20": signal.ma20,
+                "ma50": signal.ma50,
+                "ma200": signal.ma200,
+                "distance": ma_distance,
+            },
+            "rsi": {
+                "rsi14": signal.rsi14,
+                "status": "oversold" if signal.rsi14 < 30 else "overbought" if signal.rsi14 > 70 else "neutral",
+            },
+            "macd": {
+                "macd": signal.macd,
+                "signal": signal.macd_signal,
+                "histogram": signal.macd - signal.macd_signal,
+            },
+            "bollinger": {
+                "upper": signal.bollinger_upper,
+                "mid": signal.bollinger_mid,
+                "lower": signal.bollinger_lower,
+            },
+            "volatility": {
+                "vol_10d": vol_10,
+                "vol_20d": vol_20,
+                "vol_60d": vol_60,
+                "annualized": signal.annualized_volatility_pct,
+                "risk_level": signal.risk_level,
+            },
+            "trend": signal.trend,
+            "mean_reversion": signal.mean_reversion_signal,
+        }
+
+        await websocket.send(json.dumps({
+            "type": "stock_technical_indicators_loaded",
+            "ticker": ticker,
+            "indicators": indicators,
+        }, ensure_ascii=False, default=str))
+    except Exception as exc:
+        logger.exception("Error getting technical indicators for %s", ticker)
+        await websocket.send(json.dumps({
+            "type": "stock_technical_indicators_loaded",
+            "ticker": ticker,
+            "indicators": None,
+            "error": str(exc),
+        }, ensure_ascii=False))
+
+
+async def handle_run_stock_enrich(gateway: Any, websocket: Any, data: dict[str, Any]) -> None:
+    ticker = normalize_symbol(data.get("ticker", ""))
+    start_date = str(data.get("start_date") or "").strip()[:10]
+    end_date = str(data.get("end_date") or "").strip()[:10]
+    story_date = str(data.get("story_date") or end_date or "").strip()[:10]
+    target_date = str(data.get("target_date") or "").strip()[:10]
+    force = bool(data.get("force", False))
+    rebuild_story = bool(data.get("rebuild_story", True))
+    rebuild_similar_days = bool(data.get("rebuild_similar_days", True))
+    only_local_to_llm = bool(data.get("only_local_to_llm", False))
+    limit = data.get("limit", 200)
+
+    try:
+        limit = max(10, min(int(limit), 500))
+    except (TypeError, ValueError):
+        limit = 200
+
+    if not ticker or not start_date or not end_date:
+        await websocket.send(json.dumps({
+            "type": "stock_enrich_completed",
+            "ticker": ticker,
+            "start_date": start_date,
+            "end_date": end_date,
+            "error": "ticker, start_date, end_date are required",
+        }, ensure_ascii=False))
+        return
+
+    if only_local_to_llm and not llm_enrichment_enabled():
+        await websocket.send(json.dumps({
+            "type": "stock_enrich_completed",
+            "ticker": ticker,
+            "start_date": start_date,
+            "end_date": end_date,
+            "error": "only_local_to_llm requires EXPLAIN_ENRICH_USE_LLM=true and a configured LLM provider",
+        }, ensure_ascii=False))
+        return
+
+    result = await asyncio.to_thread(
+        enrich_news_for_symbol,
+        gateway.storage.market_store,
+        ticker,
+        start_date=start_date,
+        end_date=end_date,
+        limit=limit,
+        skip_existing=not force,
+        only_reanalyze_local=only_local_to_llm,
+    )
+
+    story_status = None
+    if rebuild_story and story_date:
+        await asyncio.to_thread(gateway.storage.market_store.delete_story_cache, ticker, as_of_date=story_date)
+        story_result = await asyncio.to_thread(
+            news_domain.get_story_payload,
+            gateway.storage.market_store,
+            ticker=ticker,
+            as_of_date=story_date,
+        )
+        story_status = {"as_of_date": story_date, "source": story_result.get("source") or "local"}
+
+    similar_status = None
+    if rebuild_similar_days and target_date:
+        await asyncio.to_thread(gateway.storage.market_store.delete_similar_day_cache, ticker, target_date=target_date)
+        similar_result = await asyncio.to_thread(
+            news_domain.get_similar_days_payload,
+            gateway.storage.market_store,
+            ticker=ticker,
+            date=target_date,
+            n_similar=8,
+        )
+        similar_status = {
+            "target_date": target_date,
+            "count": len(similar_result.get("items") or []),
+            "error": similar_result.get("error"),
+        }
+
+    await websocket.send(json.dumps({
+        "type": "stock_enrich_completed",
+        "ticker": ticker,
+        "start_date": start_date,
+        "end_date": end_date,
+        "story_date": story_date or None,
+        "target_date": target_date or None,
+        "force": force,
+        "only_local_to_llm": only_local_to_llm,
+        "stats": result,
+        "story_status": story_status,
+        "similar_status": similar_status,
+    }, ensure_ascii=False, default=str))
--- a/backend/services/market.py
+++ b/backend/services/market.py
@@ -54,6 +54,7 @@ class MarketService:
        self.running = False
        self._loop: Optional[asyncio.AbstractEventLoop] = None
        self._broadcast_func: Optional[Callable] = None
+        self._price_record_func: Optional[Callable[..., None]] = None
        self._price_manager: Optional[Any] = None
        self._current_date: Optional[str] = None

@@ -64,6 +65,18 @@ class MarketService:
        self._session_start_values: Optional[Dict[str, float]] = None
        self._session_start_timestamp: Optional[int] = None

+    def get_live_quote_provider(self) -> Optional[str]:
+        """Return the active live quote provider for UI/debugging."""
+        if self.backtest_mode:
+            return "backtest"
+        if self.mock_mode:
+            return "mock"
+        if self._price_manager and hasattr(self._price_manager, "provider"):
+            provider = getattr(self._price_manager, "provider", None)
+            if isinstance(provider, str) and provider.strip():
+                return provider.strip().lower()
+        return None
+
    @property
    def mode_name(self) -> str:
        if self.backtest_mode:
@@ -92,6 +105,10 @@ class MarketService:
            f"Market service started: {self.mode_name}, tickers={self.tickers}",  # noqa: E501
        )

+    def set_price_recorder(self, recorder: Optional[Callable[..., None]]):
+        """Register an optional callback for persisting runtime price points."""
+        self._price_record_func = recorder
+
    def _make_price_callback(self) -> Callable:
        """Create thread-safe price callback"""

@@ -169,6 +186,24 @@ class MarketService:
            ((price - open_price) / open_price) * 100 if open_price > 0 else 0
        )

+        if self._price_record_func:
+            try:
+                self._price_record_func(
+                    ticker=symbol,
+                    timestamp=str(price_data.get("timestamp") or datetime.now().isoformat()),
+                    price=float(price),
+                    open_price=float(open_price) if open_price is not None else None,
+                    ret=float(ret),
+                    source=self.mode_name.lower(),
+                    meta=price_data,
+                )
+            except Exception as exc:
+                logger.warning(
+                    "Failed to record price point for %s: %s",
+                    symbol,
+                    exc,
+                )
+
        await self._broadcast_func(
            {
                "type": "price_update",
@@ -205,6 +240,43 @@ class MarketService:
        self._loop = None
        self._broadcast_func = None

+    def update_tickers(self, tickers: List[str]) -> Dict[str, List[str]]:
+        """Hot-update subscribed tickers without restarting the service."""
+        normalized: List[str] = []
+        for ticker in tickers:
+            symbol = normalize_symbol(ticker)
+            if symbol and symbol not in normalized:
+                normalized.append(symbol)
+
+        previous = list(self.tickers)
+        removed = [ticker for ticker in previous if ticker not in normalized]
+        added = [ticker for ticker in normalized if ticker not in previous]
+        self.tickers = normalized
+
+        if self._price_manager:
+            if removed:
+                self._price_manager.unsubscribe(removed)
+            if added:
+                if self.mock_mode:
+                    self._price_manager.subscribe(
+                        added,
+                        base_prices={ticker: 100.0 for ticker in added},
+                    )
+                else:
+                    self._price_manager.subscribe(added)
+
+            if self.backtest_mode and self._current_date:
+                self._price_manager.set_date(self._current_date)
+
+        for ticker in removed:
+            self.cache.pop(ticker, None)
+
+        return {
+            "added": added,
+            "removed": removed,
+            "active": list(self.tickers),
+        }
+
    # Backtest methods
    def set_backtest_date(self, date: str):
        """Set current backtest date"""
@@ -472,6 +544,7 @@ class MarketService:
                "status": MarketStatus.OPEN,
                "status_text": "Backtest Mode",
                "is_trading_day": True,
+                "live_quote_provider": self.get_live_quote_provider(),
            }

        now = self._now_nyse()
@@ -484,6 +557,7 @@ class MarketService:
                "status": MarketStatus.CLOSED,
                "status_text": "Market Closed (Non-trading Day)",
                "is_trading_day": False,
+                "live_quote_provider": self.get_live_quote_provider(),
            }

        market_open, market_close = self._get_market_hours(today)
@@ -493,6 +567,7 @@ class MarketService:
                "status": MarketStatus.CLOSED,
                "status_text": "Market Closed",
                "is_trading_day": is_trading,
+                "live_quote_provider": self.get_live_quote_provider(),
            }

        # Determine status based on current time
@@ -503,6 +578,7 @@ class MarketService:
                "is_trading_day": True,
                "market_open": market_open.isoformat(),
                "market_close": market_close.isoformat(),
+                "live_quote_provider": self.get_live_quote_provider(),
            }
        elif now > market_close:
            return {
@@ -511,6 +587,7 @@ class MarketService:
                "is_trading_day": True,
                "market_open": market_open.isoformat(),
                "market_close": market_close.isoformat(),
+                "live_quote_provider": self.get_live_quote_provider(),
            }
        else:
            return {
@@ -519,6 +596,7 @@ class MarketService:
                "is_trading_day": True,
                "market_open": market_open.isoformat(),
                "market_close": market_close.isoformat(),
+                "live_quote_provider": self.get_live_quote_provider(),
            }

    async def check_and_broadcast_market_status(self):
--- a/backend/services/research_db.py
+++ b/backend/services/research_db.py
@@ -0,0 +1,280 @@
+# -*- coding: utf-8 -*-
+"""Query-oriented storage for explain/research data."""
+
+from __future__ import annotations
+
+import json
+import sqlite3
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, Iterable
+
+from shared.schema import CompanyNews
+
+
+SCHEMA = """
+CREATE TABLE IF NOT EXISTS news_items (
+    id TEXT PRIMARY KEY,
+    ticker TEXT NOT NULL,
+    published_at TEXT,
+    trade_date TEXT,
+    source TEXT,
+    title TEXT NOT NULL,
+    summary TEXT,
+    url TEXT,
+    related TEXT,
+    category TEXT,
+    raw_json TEXT NOT NULL,
+    ingest_run_date TEXT,
+    created_at TEXT NOT NULL
+);
+CREATE INDEX IF NOT EXISTS idx_news_items_ticker_date
+ON news_items (ticker, trade_date DESC, published_at DESC);
+"""
+
+
+def _json_dumps(value: Any) -> str:
+    return json.dumps(value, ensure_ascii=False, sort_keys=True, default=str)
+
+
+def _resolve_news_id(ticker: str, item: CompanyNews, fallback_index: int) -> str:
+    base = item.url or item.title or f"{ticker}-{fallback_index}"
+    return f"{ticker}:{base}"
+
+
+def _resolve_trade_date(date_value: str | None) -> str | None:
+    if not date_value:
+        return None
+    normalized = str(date_value).strip()
+    if not normalized:
+        return None
+    if "T" in normalized:
+        return normalized.split("T", 1)[0]
+    if " " in normalized:
+        return normalized.split(" ", 1)[0]
+    return normalized[:10]
+
+
+class ResearchDb:
+    """Small SQLite helper for explain-oriented news storage."""
+
+    def __init__(self, db_path: Path):
+        self.db_path = Path(db_path)
+        self.db_path.parent.mkdir(parents=True, exist_ok=True)
+        self._init_db()
+
+    def _connect(self) -> sqlite3.Connection:
+        conn = sqlite3.connect(self.db_path)
+        conn.row_factory = sqlite3.Row
+        conn.execute("PRAGMA journal_mode=WAL")
+        conn.execute("PRAGMA foreign_keys=ON")
+        return conn
+
+    def _init_db(self):
+        with self._connect() as conn:
+            conn.executescript(SCHEMA)
+
+    def upsert_news_items(
+        self,
+        *,
+        ticker: str,
+        items: Iterable[CompanyNews],
+        ingest_run_date: str | None = None,
+    ) -> list[dict[str, Any]]:
+        """Persist provider news and return normalized rows."""
+        normalized_rows: list[dict[str, Any]] = []
+        timestamp = datetime.utcnow().isoformat(timespec="seconds")
+        symbol = str(ticker or "").strip().upper()
+        if not symbol:
+            return normalized_rows
+
+        with self._connect() as conn:
+            for index, item in enumerate(items):
+                news_id = _resolve_news_id(symbol, item, index)
+                trade_date = _resolve_trade_date(item.date)
+                payload = item.model_dump()
+                row = {
+                    "id": news_id,
+                    "ticker": symbol,
+                    "published_at": item.date,
+                    "trade_date": trade_date,
+                    "source": item.source,
+                    "title": item.title,
+                    "summary": item.summary,
+                    "url": item.url,
+                    "related": item.related,
+                    "category": item.category,
+                    "raw_json": _json_dumps(payload),
+                    "ingest_run_date": ingest_run_date,
+                    "created_at": timestamp,
+                }
+                conn.execute(
+                    """
+                    INSERT INTO news_items
+                    (id, ticker, published_at, trade_date, source, title, summary, url,
+                     related, category, raw_json, ingest_run_date, created_at)
+                    VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+                    ON CONFLICT(id) DO UPDATE SET
+                        ticker = excluded.ticker,
+                        published_at = excluded.published_at,
+                        trade_date = excluded.trade_date,
+                        source = excluded.source,
+                        title = excluded.title,
+                        summary = excluded.summary,
+                        url = excluded.url,
+                        related = excluded.related,
+                        category = excluded.category,
+                        raw_json = excluded.raw_json,
+                        ingest_run_date = excluded.ingest_run_date
+                    """,
+                    (
+                        row["id"],
+                        row["ticker"],
+                        row["published_at"],
+                        row["trade_date"],
+                        row["source"],
+                        row["title"],
+                        row["summary"],
+                        row["url"],
+                        row["related"],
+                        row["category"],
+                        row["raw_json"],
+                        row["ingest_run_date"],
+                        row["created_at"],
+                    ),
+                )
+                normalized_rows.append(row)
+        return normalized_rows
+
+    def get_news_items(
+        self,
+        *,
+        ticker: str,
+        start_date: str | None = None,
+        end_date: str | None = None,
+        limit: int = 20,
+    ) -> list[dict[str, Any]]:
+        """Return normalized news rows for explain UI."""
+        symbol = str(ticker or "").strip().upper()
+        if not symbol:
+            return []
+
+        sql = """
+            SELECT id, ticker, published_at, trade_date, source, title, summary,
+                   url, related, category
+            FROM news_items
+            WHERE ticker = ?
+        """
+        params: list[Any] = [symbol]
+        if start_date:
+            sql += " AND COALESCE(trade_date, substr(published_at, 1, 10)) >= ?"
+            params.append(start_date)
+        if end_date:
+            sql += " AND COALESCE(trade_date, substr(published_at, 1, 10)) <= ?"
+            params.append(end_date)
+        sql += " ORDER BY COALESCE(published_at, trade_date) DESC LIMIT ?"
+        params.append(max(1, int(limit)))
+
+        with self._connect() as conn:
+            rows = conn.execute(sql, params).fetchall()
+
+        return [
+            {
+                "id": row["id"],
+                "ticker": row["ticker"],
+                "date": row["published_at"] or row["trade_date"],
+                "trade_date": row["trade_date"],
+                "source": row["source"],
+                "title": row["title"],
+                "summary": row["summary"],
+                "url": row["url"],
+                "related": row["related"],
+                "category": row["category"],
+            }
+            for row in rows
+        ]
+
+    def get_news_timeline(
+        self,
+        *,
+        ticker: str,
+        start_date: str | None = None,
+        end_date: str | None = None,
+    ) -> list[dict[str, Any]]:
+        """Aggregate news counts per trade date for chart markers."""
+        symbol = str(ticker or "").strip().upper()
+        if not symbol:
+            return []
+
+        sql = """
+            SELECT COALESCE(trade_date, substr(published_at, 1, 10)) AS date,
+                   COUNT(*) AS count,
+                   COUNT(DISTINCT source) AS source_count,
+                   MAX(title) AS top_title
+            FROM news_items
+            WHERE ticker = ?
+        """
+        params: list[Any] = [symbol]
+        if start_date:
+            sql += " AND COALESCE(trade_date, substr(published_at, 1, 10)) >= ?"
+            params.append(start_date)
+        if end_date:
+            sql += " AND COALESCE(trade_date, substr(published_at, 1, 10)) <= ?"
+            params.append(end_date)
+        sql += """
+            GROUP BY COALESCE(trade_date, substr(published_at, 1, 10))
+            ORDER BY date ASC
+        """
+
+        with self._connect() as conn:
+            rows = conn.execute(sql, params).fetchall()
+
+        return [
+            {
+                "date": row["date"],
+                "count": int(row["count"] or 0),
+                "source_count": int(row["source_count"] or 0),
+                "top_title": row["top_title"] or "",
+            }
+            for row in rows
+            if row["date"]
+        ]
+
+    def get_news_by_ids(
+        self,
+        *,
+        ticker: str,
+        article_ids: Iterable[str],
+    ) -> list[dict[str, Any]]:
+        """Return selected persisted news items."""
+        symbol = str(ticker or "").strip().upper()
+        ids = [str(article_id).strip() for article_id in article_ids if str(article_id).strip()]
+        if not symbol or not ids:
+            return []
+
+        placeholders = ",".join("?" for _ in ids)
+        sql = f"""
+            SELECT id, ticker, published_at, trade_date, source, title, summary,
+                   url, related, category
+            FROM news_items
+            WHERE ticker = ? AND id IN ({placeholders})
+            ORDER BY COALESCE(published_at, trade_date) DESC
+        """
+        with self._connect() as conn:
+            rows = conn.execute(sql, [symbol, *ids]).fetchall()
+
+        return [
+            {
+                "id": row["id"],
+                "ticker": row["ticker"],
+                "date": row["published_at"] or row["trade_date"],
+                "trade_date": row["trade_date"],
+                "source": row["source"],
+                "title": row["title"],
+                "summary": row["summary"],
+                "url": row["url"],
+                "related": row["related"],
+                "category": row["category"],
+            }
+            for row in rows
+        ]
--- a/backend/services/runtime_db.py
+++ b/backend/services/runtime_db.py
@@ -0,0 +1,512 @@
+# -*- coding: utf-8 -*-
+"""Run-scoped SQLite storage for query-oriented runtime history."""
+
+from __future__ import annotations
+
+import hashlib
+import json
+import sqlite3
+from pathlib import Path
+from typing import Any, Dict, Iterable, Optional
+
+
+SCHEMA = """
+CREATE TABLE IF NOT EXISTS events (
+    id TEXT PRIMARY KEY,
+    event_type TEXT NOT NULL,
+    timestamp TEXT,
+    agent_id TEXT,
+    agent_name TEXT,
+    ticker TEXT,
+    title TEXT,
+    content TEXT,
+    payload_json TEXT NOT NULL,
+    run_date TEXT
+);
+CREATE INDEX IF NOT EXISTS idx_events_type_time ON events(event_type, timestamp DESC);
+CREATE INDEX IF NOT EXISTS idx_events_ticker_time ON events(ticker, timestamp DESC);
+
+CREATE TABLE IF NOT EXISTS trades (
+    id TEXT PRIMARY KEY,
+    ticker TEXT NOT NULL,
+    side TEXT,
+    qty REAL,
+    price REAL,
+    timestamp TEXT,
+    trading_date TEXT,
+    agent_id TEXT,
+    meta_json TEXT
+);
+CREATE INDEX IF NOT EXISTS idx_trades_ticker_time ON trades(ticker, timestamp DESC);
+
+CREATE TABLE IF NOT EXISTS signals (
+    id TEXT PRIMARY KEY,
+    ticker TEXT NOT NULL,
+    agent_id TEXT,
+    agent_name TEXT,
+    role TEXT,
+    signal TEXT,
+    confidence REAL,
+    reasoning_json TEXT,
+    reasons_json TEXT,
+    risks_json TEXT,
+    invalidation TEXT,
+    next_action TEXT,
+    intrinsic_value REAL,
+    fair_value_range_json TEXT,
+    value_gap_pct REAL,
+    valuation_methods_json TEXT,
+    real_return REAL,
+    is_correct TEXT,
+    trade_date TEXT,
+    created_at TEXT,
+    meta_json TEXT
+);
+CREATE INDEX IF NOT EXISTS idx_signals_ticker_date ON signals(ticker, trade_date DESC);
+CREATE INDEX IF NOT EXISTS idx_signals_agent_date ON signals(agent_id, trade_date DESC);
+
+CREATE TABLE IF NOT EXISTS price_points (
+    id TEXT PRIMARY KEY,
+    ticker TEXT NOT NULL,
+    timestamp TEXT NOT NULL,
+    price REAL NOT NULL,
+    open_price REAL,
+    ret REAL,
+    source TEXT,
+    meta_json TEXT
+);
+CREATE INDEX IF NOT EXISTS idx_price_points_ticker_time ON price_points(ticker, timestamp DESC);
+"""
+
+
+def _json_dumps(value: Any) -> str:
+    return json.dumps(value, ensure_ascii=False, sort_keys=True, default=str)
+
+
+def _hash_key(*parts: Any) -> str:
+    raw = "::".join("" if part is None else str(part) for part in parts)
+    return hashlib.sha1(raw.encode("utf-8")).hexdigest()
+
+
+class RuntimeDb:
+    """Small SQLite helper for append-mostly runtime data."""
+
+    def __init__(self, db_path: Path):
+        self.db_path = Path(db_path)
+        self.db_path.parent.mkdir(parents=True, exist_ok=True)
+        self._init_db()
+
+    def _connect(self) -> sqlite3.Connection:
+        conn = sqlite3.connect(self.db_path)
+        conn.row_factory = sqlite3.Row
+        conn.execute("PRAGMA journal_mode=WAL")
+        conn.execute("PRAGMA foreign_keys=ON")
+        return conn
+
+    def _init_db(self):
+        with self._connect() as conn:
+            conn.executescript(SCHEMA)
+
+    def insert_event(self, event: Dict[str, Any]):
+        payload = dict(event or {})
+        if not payload:
+            return
+
+        event_id = payload.get("id") or _hash_key(
+            payload.get("type"),
+            payload.get("timestamp"),
+            payload.get("agentId") or payload.get("agent_id"),
+            payload.get("content"),
+            payload.get("title"),
+        )
+        ticker = payload.get("ticker")
+        if not ticker and isinstance(payload.get("tickers"), list) and len(payload["tickers"]) == 1:
+            ticker = payload["tickers"][0]
+
+        with self._connect() as conn:
+            conn.execute(
+                """
+                INSERT OR IGNORE INTO events
+                (id, event_type, timestamp, agent_id, agent_name, ticker, title, content, payload_json, run_date)
+                VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+                """,
+                (
+                    event_id,
+                    payload.get("type"),
+                    payload.get("timestamp"),
+                    payload.get("agentId") or payload.get("agent_id"),
+                    payload.get("agentName") or payload.get("agent_name"),
+                    ticker,
+                    payload.get("title"),
+                    payload.get("content"),
+                    _json_dumps(payload),
+                    payload.get("date") or payload.get("trading_date") or payload.get("run_date"),
+                ),
+            )
+
+    def get_recent_feed_events(
+        self,
+        *,
+        limit: int = 200,
+        event_types: Optional[Iterable[str]] = None,
+    ) -> list[Dict[str, Any]]:
+        """Return recent persisted feed events in newest-first order."""
+        event_types = tuple(event_types or ())
+        sql = """
+            SELECT payload_json
+            FROM events
+        """
+        params: list[Any] = []
+        if event_types:
+            placeholders = ",".join("?" for _ in event_types)
+            sql += f" WHERE event_type IN ({placeholders})"
+            params.extend(event_types)
+        sql += " ORDER BY timestamp DESC LIMIT ?"
+        params.append(max(1, int(limit)))
+
+        with self._connect() as conn:
+            rows = conn.execute(sql, params).fetchall()
+
+        items: list[Dict[str, Any]] = []
+        for row in rows:
+            try:
+                payload = json.loads(row["payload_json"]) if row["payload_json"] else {}
+            except json.JSONDecodeError:
+                payload = {}
+            if payload:
+                items.append(payload)
+        return items
+
+    def get_last_day_feed_events(
+        self,
+        *,
+        current_date: Optional[str] = None,
+        limit: int = 200,
+        event_types: Optional[Iterable[str]] = None,
+    ) -> list[Dict[str, Any]]:
+        """Return latest trading day events in newest-first order for replay."""
+        event_types = tuple(event_types or ())
+        target_date = str(current_date or "").strip() or None
+
+        with self._connect() as conn:
+            if not target_date:
+                row = conn.execute(
+                    """
+                    SELECT run_date
+                    FROM events
+                    WHERE run_date IS NOT NULL AND TRIM(run_date) != ''
+                    ORDER BY run_date DESC
+                    LIMIT 1
+                    """
+                ).fetchone()
+                target_date = row["run_date"] if row else None
+
+            if not target_date:
+                return []
+
+            sql = """
+                SELECT payload_json
+                FROM events
+                WHERE run_date = ?
+            """
+            params: list[Any] = [target_date]
+            if event_types:
+                placeholders = ",".join("?" for _ in event_types)
+                sql += f" AND event_type IN ({placeholders})"
+                params.extend(event_types)
+            sql += " ORDER BY timestamp DESC LIMIT ?"
+            params.append(max(1, int(limit)))
+            rows = conn.execute(sql, params).fetchall()
+
+        items: list[Dict[str, Any]] = []
+        for row in rows:
+            try:
+                payload = json.loads(row["payload_json"]) if row["payload_json"] else {}
+            except json.JSONDecodeError:
+                payload = {}
+            if payload:
+                items.append(payload)
+        return items
+
+    def upsert_trade(self, trade: Dict[str, Any]):
+        payload = dict(trade or {})
+        if not payload:
+            return
+
+        trade_id = payload.get("id") or _hash_key(
+            payload.get("ticker"),
+            payload.get("timestamp") or payload.get("ts"),
+            payload.get("side"),
+            payload.get("qty"),
+            payload.get("price"),
+        )
+        with self._connect() as conn:
+            conn.execute(
+                """
+                INSERT OR REPLACE INTO trades
+                (id, ticker, side, qty, price, timestamp, trading_date, agent_id, meta_json)
+                VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)
+                """,
+                (
+                    trade_id,
+                    payload.get("ticker"),
+                    payload.get("side"),
+                    payload.get("qty"),
+                    payload.get("price"),
+                    payload.get("timestamp") or payload.get("ts"),
+                    payload.get("trading_date"),
+                    payload.get("agentId") or payload.get("agent_id"),
+                    _json_dumps(payload),
+                ),
+            )
+
+    def upsert_signal(self, signal: Dict[str, Any], *, agent_id: str, agent_name: str, role: str):
+        payload = dict(signal or {})
+        ticker = payload.get("ticker")
+        if not ticker:
+            return
+
+        signal_id = _hash_key(
+            agent_id,
+            ticker,
+            payload.get("date"),
+            payload.get("signal"),
+            payload.get("confidence"),
+        )
+        with self._connect() as conn:
+            conn.execute(
+                """
+                INSERT OR REPLACE INTO signals
+                (id, ticker, agent_id, agent_name, role, signal, confidence, reasoning_json,
+                 reasons_json, risks_json, invalidation, next_action, intrinsic_value,
+                 fair_value_range_json, value_gap_pct, valuation_methods_json,
+                 real_return, is_correct, trade_date, created_at, meta_json)
+                VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+                """,
+                (
+                    signal_id,
+                    ticker,
+                    agent_id,
+                    agent_name,
+                    role,
+                    payload.get("signal"),
+                    payload.get("confidence"),
+                    _json_dumps(payload.get("reasoning")),
+                    _json_dumps(payload.get("reasons")),
+                    _json_dumps(payload.get("risks")),
+                    payload.get("invalidation"),
+                    payload.get("next_action"),
+                    payload.get("intrinsic_value"),
+                    _json_dumps(payload.get("fair_value_range")),
+                    payload.get("value_gap_pct"),
+                    _json_dumps(payload.get("valuation_methods")),
+                    payload.get("real_return"),
+                    None if payload.get("is_correct") is None else str(payload.get("is_correct")),
+                    payload.get("date"),
+                    payload.get("created_at") or payload.get("date"),
+                    _json_dumps(payload),
+                ),
+            )
+
+    def replace_signals_for_leaderboard(self, leaderboard: Iterable[Dict[str, Any]]):
+        with self._connect() as conn:
+            conn.execute("DELETE FROM signals")
+            for agent in leaderboard:
+                agent_id = agent.get("agentId")
+                agent_name = agent.get("name")
+                role = agent.get("role")
+                for signal in agent.get("signals", []) or []:
+                    payload = dict(signal or {})
+                    ticker = payload.get("ticker")
+                    if not ticker:
+                        continue
+                    signal_id = _hash_key(
+                        agent_id,
+                        ticker,
+                        payload.get("date"),
+                        payload.get("signal"),
+                        payload.get("confidence"),
+                    )
+                    conn.execute(
+                        """
+                        INSERT INTO signals
+                        (id, ticker, agent_id, agent_name, role, signal, confidence, reasoning_json,
+                         reasons_json, risks_json, invalidation, next_action, intrinsic_value,
+                         fair_value_range_json, value_gap_pct, valuation_methods_json,
+                         real_return, is_correct, trade_date, created_at, meta_json)
+                        VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
+                        """,
+                        (
+                            signal_id,
+                            ticker,
+                            agent_id,
+                            agent_name,
+                            role,
+                            payload.get("signal"),
+                            payload.get("confidence"),
+                            _json_dumps(payload.get("reasoning")),
+                            _json_dumps(payload.get("reasons")),
+                            _json_dumps(payload.get("risks")),
+                            payload.get("invalidation"),
+                            payload.get("next_action"),
+                            payload.get("intrinsic_value"),
+                            _json_dumps(payload.get("fair_value_range")),
+                            payload.get("value_gap_pct"),
+                            _json_dumps(payload.get("valuation_methods")),
+                            payload.get("real_return"),
+                            None if payload.get("is_correct") is None else str(payload.get("is_correct")),
+                            payload.get("date"),
+                            payload.get("created_at") or payload.get("date"),
+                            _json_dumps(payload),
+                        ),
+                    )
+
+    def insert_price_point(
+        self,
+        *,
+        ticker: str,
+        timestamp: str,
+        price: float,
+        open_price: Optional[float] = None,
+        ret: Optional[float] = None,
+        source: Optional[str] = None,
+        meta: Optional[Dict[str, Any]] = None,
+    ):
+        price_id = _hash_key(ticker, timestamp, price, open_price, ret)
+        with self._connect() as conn:
+            conn.execute(
+                """
+                INSERT OR IGNORE INTO price_points
+                (id, ticker, timestamp, price, open_price, ret, source, meta_json)
+                VALUES (?, ?, ?, ?, ?, ?, ?, ?)
+                """,
+                (
+                    price_id,
+                    ticker,
+                    timestamp,
+                    price,
+                    open_price,
+                    ret,
+                    source,
+                    _json_dumps(meta or {}),
+                ),
+            )
+
+    def get_stock_explain_snapshot(
+        self,
+        ticker: str,
+        *,
+        limit_events: int = 24,
+        limit_trades: int = 12,
+        limit_signals: int = 12,
+    ) -> Dict[str, list[Dict[str, Any]]]:
+        """Fetch query-oriented history for a single ticker."""
+        symbol = str(ticker or "").strip().upper()
+        if not symbol:
+            return {"events": [], "trades": [], "signals": []}
+
+        with self._connect() as conn:
+            trade_rows = conn.execute(
+                """
+                SELECT * FROM trades
+                WHERE ticker = ?
+                ORDER BY timestamp DESC
+                LIMIT ?
+                """,
+                (symbol, limit_trades),
+            ).fetchall()
+            signal_rows = conn.execute(
+                """
+                SELECT * FROM signals
+                WHERE ticker = ?
+                ORDER BY trade_date DESC, created_at DESC
+                LIMIT ?
+                """,
+                (symbol, limit_signals),
+            ).fetchall()
+            event_rows = conn.execute(
+                """
+                SELECT * FROM events
+                WHERE payload_json LIKE ? OR content LIKE ? OR title LIKE ? OR ticker = ?
+                ORDER BY timestamp DESC
+                LIMIT ?
+                """,
+                (f"%{symbol}%", f"%{symbol}%", f"%{symbol}%", symbol, limit_events * 3),
+            ).fetchall()
+
+        normalized_events = []
+        seen_event_ids: set[str] = set()
+        for row in event_rows:
+            payload = json.loads(row["payload_json"]) if row["payload_json"] else {}
+            content = str(row["content"] or payload.get("content") or "")
+            title = str(row["title"] or payload.get("title") or "")
+            if symbol not in f"{title} {content}".upper() and str(row["ticker"] or "").upper() != symbol:
+                continue
+            event_id = row["id"]
+            if event_id in seen_event_ids:
+                continue
+            seen_event_ids.add(event_id)
+            normalized_events.append(
+                {
+                    "id": event_id,
+                    "type": "mention",
+                    "timestamp": row["timestamp"],
+                    "title": title or f"{row['agent_name'] or '未知角色'}提及 {symbol}",
+                    "meta": payload.get("conferenceTitle")
+                    or payload.get("feedType")
+                    or row["event_type"],
+                    "body": content,
+                    "tone": "neutral",
+                    "agent": row["agent_name"] or payload.get("agentName") or payload.get("agent"),
+                },
+            )
+            if len(normalized_events) >= limit_events:
+                break
+
+        normalized_trades = [
+            {
+                "id": row["id"],
+                "type": "trade",
+                "timestamp": row["timestamp"],
+                "title": f"{row['side']} {int(row['qty'] or 0)} 股",
+                "meta": "交易执行",
+                "body": f"成交价 ${float(row['price'] or 0):.2f}",
+                "tone": "positive" if row["side"] == "LONG" else "negative" if row["side"] == "SHORT" else "neutral",
+            }
+            for row in trade_rows
+        ]
+
+        normalized_signals = [
+            {
+                "id": row["id"],
+                "type": "signal",
+                "timestamp": f"{row['trade_date']}T08:00:00" if row["trade_date"] else row["created_at"],
+                "title": f"{row['agent_name']} 给出{row['signal'] or '中性'}信号",
+                "meta": row["role"],
+                "body": (
+                    f"后验收益 {float(row['real_return']) * 100:+.2f}%"
+                    if row["real_return"] is not None
+                    else "该信号暂未完成后验评估"
+                ),
+                "tone": "positive" if str(row["signal"] or "").lower() in {"bullish", "buy", "long"} else "negative" if str(row["signal"] or "").lower() in {"bearish", "sell", "short"} else "neutral",
+                # Extended signal fields
+                "signal": row["signal"],
+                "confidence": row["confidence"],
+                "reasoning": json.loads(row["reasoning_json"]) if row["reasoning_json"] else None,
+                "reasons": json.loads(row["reasons_json"]) if row["reasons_json"] else None,
+                "risks": json.loads(row["risks_json"]) if row["risks_json"] else None,
+                "invalidation": row["invalidation"],
+                "next_action": row["next_action"],
+                "intrinsic_value": row["intrinsic_value"],
+                "fair_value_range": json.loads(row["fair_value_range_json"]) if row["fair_value_range_json"] else None,
+                "value_gap_pct": row["value_gap_pct"],
+                "valuation_methods": json.loads(row["valuation_methods_json"]) if row["valuation_methods_json"] else None,
+            }
+            for row in signal_rows
+        ]
+
+        return {
+            "events": normalized_events,
+            "trades": normalized_trades,
+            "signals": normalized_signals,
+        }
--- a/backend/services/storage.py
+++ b/backend/services/storage.py
@@ -10,6 +10,10 @@ from datetime import datetime
 from pathlib import Path
 from typing import Any, Dict, List, Optional

+from backend.data.market_store import MarketStore
+from .research_db import ResearchDb
+from .runtime_db import RuntimeDb
+
 logger = logging.getLogger(__name__)


@@ -61,6 +65,9 @@ class StorageService:
        self.state_dir = self.dashboard_dir.parent / "state"
        self.state_dir.mkdir(parents=True, exist_ok=True)
        self.server_state_file = self.state_dir / "server_state.json"
+        self.runtime_db = RuntimeDb(self.state_dir / "runtime.db")
+        self.research_db = ResearchDb(self.state_dir / "research.db")
+        self.market_store = MarketStore()

        # Feed history (for agent messages)
        self.max_feed_history = 200
@@ -114,6 +121,11 @@ class StorageService:
        try:
            with open(file_path, "w", encoding="utf-8") as f:
                json.dump(data, f, indent=2, ensure_ascii=False)
+            if file_type == "leaderboard" and isinstance(data, list):
+                self.runtime_db.replace_signals_for_leaderboard(data)
+            elif file_type == "trades" and isinstance(data, list):
+                for trade in data:
+                    self.runtime_db.upsert_trade(trade)
        except Exception as e:
            logger.error(f"Failed to save {file_type}.json: {e}")

@@ -211,6 +223,7 @@ class StorageService:
        try:
            with open(self.internal_state_file, "w", encoding="utf-8") as f:
                json.dump(state, f, indent=2, ensure_ascii=False)
+            self._sync_price_history_to_db(state.get("price_history", {}))
        except Exception as e:
            logger.error(f"Failed to save internal state: {e}")

@@ -231,6 +244,41 @@ class StorageService:
            "margin_requirement": 0.25,  # Default 25% margin requirement
        }

+    @staticmethod
+    def _portfolio_is_pristine(portfolio_state: Dict[str, Any]) -> bool:
+        """Return whether the persisted portfolio can be safely rebased."""
+        positions = portfolio_state.get("positions", {})
+        has_positions = any(
+            position.get("long", 0) or position.get("short", 0)
+            for position in positions.values()
+        )
+        margin_used = float(portfolio_state.get("margin_used", 0.0) or 0.0)
+        return not has_positions and margin_used == 0.0
+
+    def can_apply_initial_cash(self) -> bool:
+        """Only allow initial cash changes before the run has traded."""
+        state = self.load_internal_state()
+        if not self._portfolio_is_pristine(state.get("portfolio_state", {})):
+            return False
+        if state.get("all_trades"):
+            return False
+        return len(state.get("equity_history", [])) <= 1
+
+    def apply_initial_cash(self, initial_cash: float) -> bool:
+        """Rebase storage state to a new initial cash when the run is pristine."""
+        if not self.can_apply_initial_cash():
+            return False
+
+        self.initial_cash = float(initial_cash)
+        if self.internal_state_file.exists():
+            self.internal_state_file.unlink()
+
+        self.initialize_empty_dashboard()
+        state = self.load_server_state()
+        self.update_server_state_from_dashboard(state)
+        self.save_server_state(state)
+        return True
+
    def save_portfolio_state(self, portfolio: Dict[str, Any]):
        """
        Save portfolio state to internal state
@@ -750,6 +798,7 @@ class StorageService:
            "last_day_history": [],
            "trading_days_total": 0,
            "trading_days_completed": 0,
+            "price_history": {},
        }

        if not self.server_state_file.exists():
@@ -771,6 +820,11 @@ class StorageService:
        )
        logger.info(f"Trades: {len(saved_state.get('trades', []))} records")

+        for event in saved_state.get("feed_history", []):
+            self.runtime_db.insert_event(event)
+        for trade in saved_state.get("trades", []):
+            self.runtime_db.upsert_trade(trade)
+
        return saved_state

    def save_server_state(self, state: Dict[str, Any]):
@@ -852,6 +906,7 @@ class StorageService:
            state["feed_history"] = []

        state["feed_history"].insert(0, feed_msg)
+        self.runtime_db.insert_event(feed_msg)

        # Trim to max size
        if len(state["feed_history"]) > self.max_feed_history:
@@ -861,6 +916,69 @@ class StorageService:

        return True

+    def record_price_point(
+        self,
+        *,
+        ticker: str,
+        timestamp: str,
+        price: float,
+        open_price: Optional[float] = None,
+        ret: Optional[float] = None,
+        source: Optional[str] = None,
+        meta: Optional[Dict[str, Any]] = None,
+    ):
+        """Persist a runtime price point for later query-oriented reads."""
+        if not ticker or not timestamp:
+            return
+        try:
+            self.runtime_db.insert_price_point(
+                ticker=ticker,
+                timestamp=timestamp,
+                price=price,
+                open_price=open_price,
+                ret=ret,
+                source=source,
+                meta=meta,
+            )
+        except Exception as exc:
+            logger.warning("Failed to record price point for %s: %s", ticker, exc)
+
+    def _sync_price_history_to_db(self, price_history: Dict[str, Any]):
+        """Backfill structured price points from serialized internal state."""
+        if not isinstance(price_history, dict):
+            return
+        for ticker, points in price_history.items():
+            if not ticker or not isinstance(points, list):
+                continue
+            for point in points:
+                if isinstance(point, (list, tuple)) and len(point) >= 2:
+                    timestamp, price = point[0], point[1]
+                    try:
+                        self.record_price_point(
+                            ticker=str(ticker),
+                            timestamp=str(timestamp),
+                            price=float(price),
+                        )
+                    except (TypeError, ValueError):
+                        continue
+                elif isinstance(point, dict):
+                    timestamp = point.get("timestamp") or point.get("label") or point.get("date")
+                    price = point.get("price") or point.get("close") or point.get("value")
+                    if not timestamp or price is None:
+                        continue
+                    try:
+                        self.record_price_point(
+                            ticker=str(ticker),
+                            timestamp=str(timestamp),
+                            price=float(price),
+                            open_price=point.get("open"),
+                            ret=point.get("ret"),
+                            source=point.get("source"),
+                            meta=point,
+                        )
+                    except (TypeError, ValueError):
+                        continue
+
    def _get_default_stats(self) -> Dict[str, Any]:
        """Get default stats structure"""
        return {
@@ -889,6 +1007,7 @@ class StorageService:
        stats = self.load_file("stats") or self._get_default_stats()
        trades = self.load_file("trades") or []
        leaderboard = self.load_file("leaderboard") or []
+        internal_state = self.load_internal_state()

        # Update state
        state["portfolio"] = {
@@ -910,6 +1029,9 @@ class StorageService:
        state["stats"] = stats
        state["trades"] = trades
        state["leaderboard"] = leaderboard
+        state["price_history"] = internal_state.get("price_history", {})
+        self.runtime_db.replace_signals_for_leaderboard(leaderboard)
+        self._sync_price_history_to_db(state["price_history"])

    # ========== Live Returns Tracking ==========

--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
cillin	3926a6bd07	feat: 架构修复 - P0/P1 问题全面修复 P0 修复: - runtimeStore: 添加缺失的 lastDayHistory 字段 - Gateway/RuntimeService: 状态同步改为内存优先，消除 glob 竞态 - App.jsx: 从 3075 行重构到 ~500 行，提取 8 个独立文件 P1 修复: - CORS: 4 个服务改为从环境变量读取允许 origins - MarketStore: 改为模块级单例模式 - Domain 层: 删除 trading thin wrapper，保留 news 真实逻辑 - 测试: 补齐 77 个 gateway/runtime 测试新增文件: - backend/tests/test_gateway.py (43 tests) - frontend/src/hooks/useWebSocketHandler.js - frontend/src/hooks/useStockRequestCallbacks.js - frontend/src/hooks/useAgentCallbacks.js - frontend/src/hooks/useRuntimeCallbacks.js - frontend/src/hooks/useWatchlistCallbacks.js - frontend/src/components/TickerBar.jsx - frontend/src/components/HeaderRight.jsx - frontend/src/components/ChartTabs.jsx Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-23 18:45:57 +08:00
cillin	80256a4079	fix(frontend): 添加缺失的 lastDayHistory 字段到 runtimeStore 修复 App.jsx 中使用不存在的 store 字段导致的潜在运行时错误。 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-23 18:20:29 +08:00
cillin	456748b01e	fix(frontend): 修复 now 变量重复声明导致的空白页面 - 从 runtimeStore 移除 now（应从 uiStore 获取） - 修复重复声明错误 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-23 18:12:32 +08:00
cillin	609b509446	feat(frontend): 完成 Zustand 状态管理迁移 - 将 App.jsx 中的 useState 迁移到 5 个 Zustand stores - useRuntimeStore: 连接状态、运行时配置 - useMarketStore: 市场数据、股票价格 - usePortfolioStore: 组合、持仓、交易 - useAgentStore: Agent 技能，工作区 - useUIStore: UI 状态、视图切换 - 保留 tickers useState（需与 INITIAL_TICKERS 同步） - 恢复 newsApi.js 和 tradingApi.js Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-23 18:03:23 +08:00
cillin	38102d0805	feat(backend): Agent 系统和 API 增强 - task_delegator: 完善团队任务分发逻辑 - runtime API: 增强运行时管理功能 - skills_manager: 技能管理改进 - tool_guard: 工具调用守卫优化 - evo_agent: 核心 Agent 改进 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-23 17:46:06 +08:00
cillin	3448667b79	feat: 微服务架构拆分和前后端优化后端: - 拆分出 agent_service, runtime_service, trading_service, news_service - Gateway 模块化拆分 (gateway_*.py) - 添加 domains/ 领域层 - 新增 control_client, runtime_client - 更新 start-dev.sh 支持 split 服务模式前端: - 完善 API 服务层 (newsApi, tradingApi) - 更新 vite.config.js - Explain 组件优化测试: - 添加多个服务 app 测试 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-23 17:45:39 +08:00
cillin	0f1bc2bb39	feat(frontend): 添加 Zustand store 架构 - 创建 5 个领域 store：runtimeStore, marketStore, portfolioStore, agentStore, uiStore - 更新 CLAUDE.md 记录架构改进 - Zustand 已安装但 stores 尚未在 App.jsx 中使用（渐进迁移） Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-23 17:44:17 +08:00
cillin	06a23c32a4	refactor: Fix code quality issues identified in analysis 1. Rename factory.py's EvoAgent data class to AgentConfig - Avoids naming conflict with base/evo_agent.py's EvoAgent 2. Export pipeline_runner functions in backend/core/__init__.py - Add create_agents, create_long_term_memory, stop_gateway 3. Consolidate PromptLoader to singleton pattern - Add get_prompt_loader() singleton function - Update all usages to use singleton Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-20 01:07:53 +08:00
cillin	5b925fbe02	feat: Refactor services architecture and update project structure - Remove Docker-based microservices (docker-compose.yml, Makefile, Dockerfiles) - Update start-dev.sh to use backend.app:app entry point - Add shared schema and client modules for service communication - Add team coordination modules (messenger, registry, task_delegator, coordinator) - Add evaluation hooks and skill adaptation hooks - Add skill template and gateway server - Update frontend WebSocket URL configuration - Add explain components for insider and technical analysis Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-20 00:57:09 +08:00
cillin	4b5ac86b83	feat: Add evaluation hooks, skill adaptation and team pipeline config - Add EvaluationHook for post-execution agent evaluation - Add SkillAdaptationHook for dynamic skill adaptation - Add team/ directory with team coordination logic - Add TEAM_PIPELINE.yaml for smoke_fullstack pipeline config - Update RuntimeView, TraderView and RuntimeSettingsPanel UI - Add runtimeApi and websocket services - Add runtime_state.json to smoke_fullstack state Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-19 18:52:12 +08:00
cillin	f4a2b7f3af	fix: Initialize watchlist draft when opening runtime settings panel - Add initialization of watchlistDraftSymbols when opening settings - Previously watchlist was empty because initial state was never set - Now copies from runtimeWatchlistSymbols on panel open Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-17 18:40:02 +08:00
cillin	2dcda63394	feat: Integrate WebSocket Gateway into API-based task launch - Gateway now starts automatically when calling POST /runtime/start - No need to run python backend/main.py separately - Single service architecture: only FastAPI (port 8000) needed - Gateway runs in background task and stops with pipeline - Add error handling and logging for Gateway lifecycle Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-17 18:36:33 +08:00
cillin	a3f767126f	feat: Complete pipeline runner integration for API-based task launch - Add backend/core/pipeline_runner.py with full pipeline execution logic - Integrate main.py pipeline startup into REST API - Add comprehensive logging and error handling - Support mock/live/backtest modes via API Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-17 18:22:22 +08:00
cillin	9ec4a8702d	feat: Add task launcher API with timestamp-based run directories - Add POST /runtime/start, /stop, /restart APIs - Run directories use timestamp format: YYYYMMDD_HHMMSS - Add force stop support in TradingRuntimeManager - Frontend: use REST API to start runtime instead of WebSocket - Remove RuntimeView page from navigation Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-17 17:14:22 +08:00
cillin	3174734f26	feat: Add runtime config validation for cash, margin and memory settings - Add initial_cash validation (must be positive number) - Add margin_requirement validation (must be non-negative) - Add enable_memory boolean configuration Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-17 16:50:47 +08:00
cillin	59b44545d0	feat: Add agent workspace system and runtime management - Add agent core modules (agent_core, factory, registry, skill_loader) - Add runtime system for agent execution management - Add REST API for agents, workspaces, and runtime control - Add process supervisor for agent lifecycle management - Add workspace template system with agent profiles - Add frontend RuntimeView and runtime API integration - Add per-agent skill workspaces for smoke_fullstack run - Refactor skill system with active/installed separation Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>	2026-03-17 16:43:29 +08:00
cillin	2daf5717ba	Add per-agent skill workspaces and TraderView management	2026-03-17 13:55:14 +08:00
cillin	1f5ee3698e	Add explain analysis workflow and UI	2026-03-16 22:28:41 +08:00
cillin	3a5558b576	Read feed replay history from runtime db	2026-03-16 02:26:24 +08:00
cillin	a41cd705b4	Prefer SQLite signals in stock explain view	2026-03-16 02:22:59 +08:00
cillin	564c92c0c8	确认PokieTicker新闻库数据源	2026-03-16 02:19:25 +08:00
cillin	78f133617f	Add run-scoped skill and prompt asset management	2026-03-16 00:04:04 +08:00
				`@@ -0,0 +1,2 @@`
				`"""News enrichment utilities for explain-oriented market research."""`