RagDocs MCP Server

此包由 MCPFlow 打包并发布到npm仓库。

基于 RAG 的文档搜索和管理的 MCP 服务器

安装与使用

直接使用npx运行:

npx @mcpflow.io/mcp-ragdocs-mcp-server

或者先安装后使用:

# 安装
npm install @mcpflow.io/mcp-ragdocs-mcp-server

# 使用
npx @mcpflow.io/mcp-ragdocs-mcp-server

使用方法

Installation

npm install -g @mcpservers/ragdocs

工具函数

initBrowser

Initializes the browser for web scraping operations.

参数:

• none: No parameters required.

cleanup

Closes the browser instance.

参数:

• none: No parameters required.

getEmbeddings

Generates embeddings for a given text.

参数:

• text: The text for which embeddings are to be generated.

initCollection

Initializes the Qdrant collection with the specified name and vector size.

参数:

• collectionName: The name of the Qdrant collection to initialize.

createCollection

Creates a new Qdrant collection with the specified name and vector size.

参数:

• vectorSize: The size of the vectors to be used in the collection.
• collectionName: The name of the Qdrant collection to create.

recreateCollection

Deletes an existing Qdrant collection and creates a new one with the specified name and vector size.

参数:

• vectorSize: The size of the vectors to be used in the collection.
• collectionName: The name of the Qdrant collection to recreate.

isHealthy

Checks if the Qdrant client is healthy.

参数:

• none: No parameters required.

addDocument

Adds a document to the Qdrant collection.

参数:

• doc: The document to add.

deleteDocument

Deletes a document from the Qdrant collection.

参数:

• url: The URL of the document to delete.

generateEmbeddings

Generates embeddings for a given text.

参数:

• text: The text for which embeddings are to be generated.

addDocumentation

Adds a document to the RAG system.

参数:

• url: URL of the document to add.
• content: Content of the document to add.
• metadata: Metadata of the document.

fetchContent

Fetches and processes content from a URL.

参数:

• url: URL to fetch content from.

chunkText

Chunks a given text into smaller pieces.

参数:

• text: The text to chunk.
• maxChunkSize: The maximum size of each chunk.

generateEmbeddings

Generates embeddings for text chunks.

参数:

• chunks: Array of text chunks.

原始信息

• 开发者: heltonteixeira
• 版本: 1.0.0
• 许可证: Apache License 2.0
• 原始仓库: heltonteixeira/ragdocs

原始README

RagDocs MCP Server

A Model Context Protocol (MCP) server that provides RAG (Retrieval-Augmented Generation) capabilities using Qdrant vector database and Ollama/OpenAI embeddings. This server enables semantic search and management of documentation through vector similarity.

Features

• Add documentation with metadata
• Semantic search through documents
• List and organize documentation
• Delete documents
• Support for both Ollama (free) and OpenAI (paid) embeddings
• Automatic text chunking and embedding generation
• Vector storage with Qdrant

Prerequisites

• Node.js 16 or higher
• One of the following Qdrant setups:
  • Local instance using Docker (free)
  • Qdrant Cloud account with API key (managed service)
• One of the following for embeddings:
  • Ollama running locally (default, free)
  • OpenAI API key (optional, paid)

Available Tools

1. add_document

Add a document to the RAG system.

Parameters:

• url (required): Document URL/identifier
• content (required): Document content
• metadata (optional): Document metadata
  • title: Document title
  • contentType: Content type (e.g., "text/markdown")

2. search_documents

Search through stored documents using semantic similarity.

Parameters:

• query (required): Natural language search query
• options (optional):
  • limit: Maximum number of results (1-20, default: 5)
  • scoreThreshold: Minimum similarity score (0-1, default: 0.7)
  • filters:
    • domain: Filter by domain
    • hasCode: Filter for documents containing code
    • after: Filter for documents after date (ISO format)
    • before: Filter for documents before date (ISO format)

3. list_documents

List all stored documents with pagination and grouping options.

Parameters (all optional):

• page: Page number (default: 1)
• pageSize: Number of documents per page (1-100, default: 20)
• groupByDomain: Group documents by domain (default: false)
• sortBy: Sort field ("timestamp", "title", or "domain")
• sortOrder: Sort order ("asc" or "desc")

4. delete_document

Delete a document from the RAG system.

Parameters:

• url (required): URL of the document to delete

Installation

npm install -g @mcpservers/ragdocs

MCP Server Configuration

{
  "mcpServers": {
    "ragdocs": {
      "command": "node",
      "args": ["@mcpservers/ragdocs"],
      "env": {
        "QDRANT_URL": "http://127.0.0.1:6333",
        "EMBEDDING_PROVIDER": "ollama"
      }
    }
  }
}

Using Qdrant Cloud:

{
  "mcpServers": {
    "ragdocs": {
      "command": "node",
      "args": ["@mcpservers/ragdocs"],
      "env": {
        "QDRANT_URL": "https://your-cluster-url.qdrant.tech",
        "QDRANT_API_KEY": "your-qdrant-api-key",
        "EMBEDDING_PROVIDER": "ollama"
      }
    }
  }
}

Using OpenAI:

{
  "mcpServers": {
    "ragdocs": {
      "command": "node",
      "args": ["@mcpservers/ragdocs"],
      "env": {
        "QDRANT_URL": "http://127.0.0.1:6333",
        "EMBEDDING_PROVIDER": "openai",
        "OPENAI_API_KEY": "your-api-key"
      }
    }
  }
}

Local Qdrant with Docker

docker run -d --name qdrant -p 6333:6333 -p 6334:6334 qdrant/qdrant

Environment Variables

• QDRANT_URL: URL of your Qdrant instance
  • For local: "http://127.0.0.1:6333" (default)
  • For cloud: "https://your-cluster-url.qdrant.tech"
• QDRANT_API_KEY: API key for Qdrant Cloud (required when using cloud instance)
• EMBEDDING_PROVIDER: Choice of embedding provider ("ollama" or "openai", default: "ollama")
• OPENAI_API_KEY: OpenAI API key (required if using OpenAI)
• EMBEDDING_MODEL: Model to use for embeddings
  • For Ollama: defaults to "nomic-embed-text"
  • For OpenAI: defaults to "text-embedding-3-small"

License

Apache License 2.0