mcp-documentation-server

概述

MCP 文档服务器是一款用于本地文档管理的先进工具，具备高级搜索功能。它使用嵌入式的 Orama 向量数据库进行混合全文和向量相似性搜索，并通过智能父子分块技术保持上下文完整性。服务器包含内置的 Web 界面用于文档管理，并通过 Google Gemini 集成提供可选的 AI 驱动的搜索功能。所有数据都存储在本地的 ~/.mcp-documentation-server/ 目录中，无需外部依赖。

试试问 AI

装完之后，这里有 5 个你可以让 AI 做的事：

• add_document

  Add a document (title, content, optional metadata)

  试试:Add a new document about React best practices试试:Create a document titled 'Project Guidelines' with team policies
• list_documents

  List all documents with metadata and content preview

  试试:Show me all documents in the knowledge base试试:List all documents with their titles and creation dates
• get_document

  Retrieve the full content of a document by ID

  试试:Get the full content of document 123试试:Retrieve the document with ID abc-123
• delete_document

  Remove a document, its chunks, database entries, and associated files

  试试:Delete document with ID 456试试:Remove the outdated marketing strategy document
• process_uploads

  Process all files in the uploads folder (chunking + embeddings)

  试试:Process all uploaded files试试:Handle the new documents in the uploads folder
• get_uploads_path

  Returns the absolute path to the uploads folder

  试试:Where are the uploaded files stored?试试:Show me the path to the uploads directory
• list_uploads_files

  Lists files in the uploads folder with size and format info

  试试:What files are in the uploads folder?试试:Show me all pending uploads with their sizes
• get_ui_url

  Returns the Web UI URL (e.g. http://localhost:3080)

  试试:What's the URL for the documentation dashboard?试试:Open the web interface
• search_documents

  Semantic vector search within a specific document

  试试:Search for 'authentication methods' in document 789试试:Find all references to API keys in the technical doc
• search_all_documents

  Hybrid (full-text + vector) cross-document search

  试试:Search across all documents for 'performance optimization'试试:Find information about database migrations in the knowledge base
• get_context_window

  Returns a window of chunks around a given chunk index

  试试:Show me context around chunk 123试试:Get neighboring chunks for position 456
• search_documents_with_ai

  AI-powered search using Gemini (requires GEMINI_API_KEY)

  试试:Use AI to analyze all documents about user feedback试试:Find insights about recent feature requests using AI

可对比工具

安装

{
  "mcpServers": {
    "documentation": {
      "command": "npx",
      "args": [
        "-y",
        "@andrea9293/mcp-documentation-server"
      ],
      "env": {
        "MCP_BASE_DIR": "/path/to/workspace",
        "GEMINI_API_KEY": "your-api-key-here",
        "MCP_EMBEDDING_MODEL": "Xenova/all-MiniLM-L6-v2",
        "START_WEB_UI": "true",
        "WEB_PORT": "3080"
      }
    }
  }
}

npm install -g @andrea9293/mcp-documentation-server

FAQ

服务器支持哪些文件格式？

服务器支持 .txt、.md 和 .pdf 文件的上传和处理。

Web UI 是可选的吗？

是的，您可以通过在环境配置中设置 START_WEB_UI=false 来禁用 Web UI。

mcp-documentation-server 对比