RAG Documentation

🚀 RAG Document MCP Server

An MCP server implementation that provides tools for retrieving and processing documents through vector search, enabling AI assistants to enhance their responses with relevant document context.

Table of Contents

• ✨ Features
• 🚀 Quick Start
• 📦 Docker Compose Setup
• 🌐 Web Interface
• ⚙️ Configuration
  • Cline Configuration
  • Claude Desktop Configuration
• 📋 Default Configuration
• 🙏 Acknowledgments
• 🔧 Troubleshooting

✨ Features

This MCP server offers the following features:

1. Document Retrieval: Efficiently retrieve relevant documents through vector search.
2. Context Enhancement: Enable AI assistants to reference and understand document content, improving response quality.
3. Multilingual Support: Support document processing and retrieval in multiple languages.

🚀 Quick Start

Install Dependencies

Make sure the following tools are installed:

• Node.js (recommended version 14+)
• Docker (optional, for Docker Compose)

Clone the Repository

git clone https://github.com/your-repository.git
cd mcp-ragdocs

Run the Server

npm install
node index.js

By default, the server will run on http://localhost:3030.

📦 Docker Compose Setup

Use the following docker-compose.yml file to start the service:

version: '3'
services:
  mcp-ragdocs:
    build: .
    ports:
      - "3030:3030"
    environment:
      - EMBEDDING_PROVIDER=ollama
      - OPENAI_API_KEY=your-api-key-here

Run the command:

docker-compose up --build

🌐 Web Interface

Visit http://localhost:3030 to use the default web interface for document retrieval and management.

⚙️ Configuration

Cline Configuration

Set the following environment variables in the .env file:

EMBEDDING_PROVIDER=ollama  # Optional values: ollama or openai
OPENAI_API_KEY=your-api-key-here  # Required (if using OpenAI)
QDRANT_URL=http://localhost:6333  # Optional, default is http://localhost:6333

Claude Desktop Configuration

Set the following content in the config.json file:

{
  "EMBEDDING_PROVIDER": "ollama",
  "OPENAI_API_KEY": "your-api-key-here",
  "QDRANT_URL": "http://localhost:6333"
}

📋 Default Configuration

The system uses Ollama as the embedding provider by default. For optimal reliability:

1. Install and run Ollama locally.
2. Configure OpenAI as a fallback (recommended):

{
  "EMBEDDING_MODEL": "nomic-embed-text", // Optional
  "FALLBACK_PROVIDER": "openai",
  "FALLBACK_MODEL": "text-embedding-3-small",
  "OPENAI_API_KEY": "your-api-key-here"
}

This configuration ensures:

• Use Ollama for fast local embedding generation.
• Automatically fallback to OpenAI if Ollama fails.
• Make external API calls only when necessary.

Note: The system will automatically use the appropriate vector dimensions based on the provider:

• Ollama (nomic-embed-text): 768 dimensions
• OpenAI (text-embedding-3-small): 1536 dimensions

🙏 Acknowledgments

This project is a fork of qpd-v/mcp-ragdocs, originally developed by qpd-v. The original project provided the foundation for this implementation. Special thanks to the original author qpd-v for their innovative work on the initial version. This fork has been improved by Rahul Retnan, adding more features.

🔧 Troubleshooting

Server Fails to Start (Port Conflict)

If the MCP server fails to start due to a port conflict, follow these steps:

1. Identify and terminate the process using port 3030:

npx kill-port 3030

2. Restart the MCP server.

3. If the problem persists, check if any other processes are using the port:

lsof -i :3030

4. If necessary, change the default port in the configuration.