Skip to main content

基于FastAPI的本地知识库MCP服务器,集成FAISS向量库

Project description

zhdddd-mcp-knowledge-server

基于FastAPI的本地知识库MCP服务器,集成FAISS向量库,支持文档上传、向量检索和MCP协议接口。

功能特性

  • 文档上传与解析:支持上传文档并自动拆分为文本块,提高检索精度
  • 向量检索:使用FAISS向量库实现高效的相似度搜索
  • MCP协议接口:支持标准的MCP协议,供大模型调用
  • 工具列表与调用:实现了/tools/list/tools/call接口,符合MCP规范
  • 健康检查:提供服务健康状态检查接口

安装方法

从PyPI安装

pip install zhdddd-mcp-knowledge-server

使用uvx启动

uvx zhdddd-mcp-knowledge-server

快速开始

  1. 启动服务器

    zhdddd-mcp-knowledge-server
    

    或使用uvx:

    uvx zhdddd-mcp-knowledge-server
    
  2. 服务默认运行在

    http://0.0.0.0:8000
    

主要接口

健康检查

  • GET /health - 检查服务状态

文档上传

  • POST /upload/document - 上传文档并解析入库

向量检索

  • POST /retrieve - 检索相似文档

MCP协议接口

  • POST /mcp/invoke - MCP协议调用接口
  • GET /mcp/info - 服务信息和工具列表

工具接口(标准MCP规范)

  • GET /tools/list - 获取可用工具列表
  • POST /tools/call - 调用指定工具

集合信息

  • GET /collection/info - 获取知识库信息

示例使用

上传文档

curl -X POST "http://localhost:8000/upload/document" \
  -F "file=@example.txt"

检索文档

curl -X POST "http://localhost:8000/retrieve" \
  -H "Content-Type: application/json" \
  -d '{"query": "ESOP是什么"}'

MCP工具调用

curl -X POST "http://localhost:8000/tools/call" \
  -H "Content-Type: application/json" \
  -d '{"tool": "retrieve_knowledge", "parameters": {"query": "ESOP是什么"}}'

配置说明

服务默认配置:

  • 主机:0.0.0.0
  • 端口:8000
  • 向量维度:768
  • 文档拆分大小:500字符
  • 文本块重叠:100字符

依赖项

  • fastapi
  • uvicorn
  • faiss-cpu
  • numpy
  • packaging
  • python-multipart
  • pydantic
  • pydantic-settings
  • python-dotenv

许可证

MIT License

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

zhdddd_mcp_knowledge_server-0.1.1.tar.gz (9.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

zhdddd_mcp_knowledge_server-0.1.1-py3-none-any.whl (8.5 kB view details)

Uploaded Python 3

File details

Details for the file zhdddd_mcp_knowledge_server-0.1.1.tar.gz.

File metadata

File hashes

Hashes for zhdddd_mcp_knowledge_server-0.1.1.tar.gz
Algorithm Hash digest
SHA256 72a54f2cde84ee3d63f2fff7a83989bb7b8d23c13f530b913d8a9ad3c705ea48
MD5 8e528e1adf831589f51292bd674588c7
BLAKE2b-256 36a444639466a19676487bd885ea241425125f407d82b9db1322c28a8ff357af

See more details on using hashes here.

File details

Details for the file zhdddd_mcp_knowledge_server-0.1.1-py3-none-any.whl.

File metadata

File hashes

Hashes for zhdddd_mcp_knowledge_server-0.1.1-py3-none-any.whl
Algorithm Hash digest
SHA256 0fef25798b83b2edf4d0852b7a7dcdc2ad69312d2246686e543e7255f641a99b
MD5 752f2fe01d1ab3bd2a613ddf5187504e
BLAKE2b-256 60848b40b5888a98ceaaa3b9d5cc49ebc31754d4704b34d0ec9478bb5caaf80f

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page