{"$schema":"https://wellknown.network/schemas/agent-record-v1.json","schemaVersion":"1","id":"ag_fe7v8jz6rz9j","handle":"iflow-mcp-kerwinmai-multi-agent-exp","url":"https://wellknown.network/agents/iflow-mcp-kerwinmai-multi-agent-exp","links":{"self":"https://wellknown.network/agents/iflow-mcp-kerwinmai-multi-agent-exp/record.json","html":"https://wellknown.network/agents/iflow-mcp-kerwinmai-multi-agent-exp","markdown":"https://wellknown.network/agents/iflow-mcp-kerwinmai-multi-agent-exp/record.md","api":"https://wellknown.network/api/v1/agents/iflow-mcp-kerwinmai-multi-agent-exp","status":"https://wellknown.network/api/v1/agents/iflow-mcp-kerwinmai-multi-agent-exp/status","claim":"https://wellknown.network/agents/iflow-mcp-kerwinmai-multi-agent-exp/claim","claimApi":"https://wellknown.network/api/v1/claims","claimDescriptor":"https://wellknown.network/agents/iflow-mcp-kerwinmai-multi-agent-exp/claim.json","badge":"https://wellknown.network/agents/iflow-mcp-kerwinmai-multi-agent-exp/badge.svg","openapi":"https://wellknown.network/openapi.json","history":"https://wellknown.network/api/v1/agents/iflow-mcp-kerwinmai-multi-agent-exp/history","tools":"https://wellknown.network/api/v1/agents/iflow-mcp-kerwinmai-multi-agent-exp/tools"},"ard":{"identifier":"urn:air::server:iflow-mcp-kerwinmai-multi-agent-exp","type":"application/mcp-server-card+json"},"kind":"mcp_server","declared":{"name":"iflow-mcp_kerwinmai-multi-agent-exp","summary":"Multi-Agent SQL Query System with MCP Support","description":"# 多智能体数据查询系统 v2.1\n\n> v2.1 新增：内置 Web 前端（Markdown 渲染、即时加载长期记忆）、REST API、启动脚本等。\n\n基于LangGraph的一主两从多智能体架构，实现智能的数据库查询和分析功能，支持长短期记忆。\n[https://github.com/user-attachments/assets/3b1b1513-3c92-4952-b8e1-7e62e77630c9](https://github.com/user-attachments/assets/3b1b1513-3c92-4952-b8e1-7e62e77630c9)\n\n## 架构设计\n\n```\n主智能体 (MasterAgent)\n├── 意图识别：判断用户问题类型（结合用户偏好）\n├── 任务路由：选择合适的处理路径\n├── 短期记忆：智能压缩对话历史（LLM总结）\n├── 长期记忆：跨会话用户偏好和知识\n└── 结果汇总：整合子智能体输出\n\n子智能体1 (SQLQueryAgent)\n└── NL2SQL转换、执行查询\n\n子智能体2 (DataAnalysisAgent)\n└── 数据分析、生成洞察\n```\n\n## 核心功能\n\n**1. 意图识别** - 自动判断4种问题类型\n- `simple_answer`: 简单问候\n- `sql_only`: 数据查询\n- `analysis_only`: 分析历史数据\n- `sql_and_analysis`: 查询+分析\n\n**2. 智能路由** - 根据意图调用不同智能体\n- 主智能体直接回答 / SQL子智能体 / 数据分析子智能体 / 两者协作\n\n**3. 双层记忆系统** ⭐NEW\n- **短期记忆（MemorySaver）**\n  - 会话内对话历史保留\n  - 智能压缩：消息>10条或token>1000时LLM自动总结\n  - 支持引用历史查询结果\n- **长期记忆（LongTermMemory）**\n  - 跨会话持久化用户偏好和知识\n  - 自动提取：对话≥4条消息时自动提取\n  - 个性化上下文：意图识别时注入用户历史\n  - SQLite存储：users、user_preferences、user_knowledge\n\n**4. 用户会话管理** ⭐NEW\n- 用户登录系统（user_id）\n- 会话隔离（session_id）\n- 智能thread_id生成（user_id_session_id）\n- 特殊命令：`new`（新会话）、`info`（查看信息）\n\n**5. MCP集成** - 通过MCP协议执行SQL\n- 独立的MCP服务器处理数据库操作\n- 支持扩展多数据源\n\n## 快速开始\n\n### 1. 安装依赖\n```bash\npip install -r requirements.txt\n```\n\n### 2. 设置API密钥\n```bash\n# Windows\nset DASHSCOPE_API_KEY=your_api_key\n\n# Linux/Mac\nexport DASHSCOPE_API_KEY=your_api_key\n```\n\n### 3. 初始化数据库\n```bash\ncd intelligent_data_query/data\npython init_db.py              # 初始化业务数据库\npython init_memory_db.py       # 初始化长期记忆数据库\n```\n\n### 4. 运行系统（命令行）\n```bash\ncd intelligent_data_query\npython agent.py\n```\n\n系统会提示输入用户ID，输入后即可开始使用。同一user_id可以在不同会话中保留个人偏好。\n\n### 5. 启动 Web 前端（v2.1）\n\n```bash\n# Windows\ncd intelligent_data_query\nstart_web.bat\n\n# Linux/Mac\ncd intelligent_data_query\nchmod +x start_web.sh\n./start_web.sh\n```\n\n浏览器访问：`http://localhost:5000`\n\n前端特性（v2.1）：\n- Web 聊天界面，支持 Markdown 渲染与代码高亮\n- 聊天区滚动优化，输入框固定底部\n- 登录后直接从 `long_term_memory.db` 读取并展示用户信息（无需对话总结）\n- REST API：`/api/login`、`/api/query`、`/api/new_session`、`/api/user_info`、`/api/health`\n\n**特殊命令…","publisher":null,"homepage":"https://github.com/KerwinMai/Multi-Agent-Exp","repository":"https://github.com/KerwinMai/Multi-Agent-Exp","version":"2.1.0","license":null,"protocols":["mcp"],"tags":["mcp"],"pricing":null,"endpoints":[{"url":"pypi:iflow-mcp_kerwinmai-multi-agent-exp","type":"package_pypi","auth":null,"probeable":false}],"skills":null,"tools":null,"extra":null,"attribution":{"kind":"pypi","name":"pypi","repoUrl":"pypi","summary":"pypi","version":"pypi","description":"pypi","homepageUrl":"pypi"}},"derived":{"capabilities":[{"slug":"data.database","name":"Databases","confidence":1,"provenance":"derived"},{"slug":"automation.orchestration","name":"Agent Orchestration","confidence":1,"provenance":"derived"},{"slug":"knowledge.memory","name":"Agent Memory","confidence":0.848,"provenance":"derived"}],"categories":["automation","data","knowledge"],"language":"en"},"observed":{"status":"unknown","statusReason":"Distributed as a package to run locally; no network endpoint to check.","lastOkAt":null,"lastProbedAt":null,"statusComputedAt":null,"reliability30d":null,"latestObservations":[],"tools":null,"package":{"name":"iflow-mcp_kerwinmai-multi-agent-exp","registry":"pypi","observedAt":"2026-10-28T05:23:55.580Z","publishedAt":"2026-02-11T12:57:49.618572Z","latestVersion":"2.1.0"},"toolSurface":null,"endpointFacts":[]},"verification":{"claimed":false,"claimedAt":null,"proofs":[]},"provenance":{"sources":[{"source":"pypi","key":"iflow-mcp_kerwinmai-multi-agent-exp","url":"https://pypi.org/project/iflow-mcp_kerwinmai-multi-agent-exp/","firstSeenAt":"2026-09-09T19:24:29.646Z","fetchedAt":"2026-10-28T05:21:53.588Z","normalizedAt":"2026-10-28T05:21:53.588Z"}]},"firstSeenAt":"2026-09-09T19:24:29.646Z","updatedAt":"2026-10-28T05:23:55.580Z"}