RAG Docs

🚀 MCP Document Retrieval Server

The MCP Document Retrieval Server is a powerful tool that enables users to efficiently manage and search through documentation. It integrates with various services and offers a range of useful tools for document handling.

🚀 Quick Start

📦 Installation

Clone the repository

git clone https://github.com/qpd-v/mcp-server-ragdocs.git
cd mcp-server-ragdocs

Install dependencies

npm install

Build the project

npm run build

✨ Features

🛠️ Available Tools

1. add_documentation

   • Add documentation from a specified URL to the RAG database.
   • Parameters:
     • url: The URL of the documentation to fetch.

2. search_documentation

   • Search through the stored documentation.
   • Parameters:
     • query: The search query.
     • limit (optional): The maximum number of results to return (default is 5).

3. list_sources

   • List all currently stored documentation sources.
   • No parameters required.

📦 Installation

Clone the repository

git clone https://github.com/qpd-v/mcp-server-ragdocs.git
cd mcp-server-ragdocs

Install dependencies

npm install

Build the project

npm run build

📚 Documentation

Configuration Guide

Environment Variable Configuration

Qdrant Service Configuration

• QDRANT_URL (Required): The URL address of the Qdrant instance.
  • Use http://localhost:6333 when running locally.
  • Use https://your-cluster-url.qdrant.tech when using cloud services.
• QDRANT_API_KEY (Required for cloud services): Your Qdrant Cloud API key.

Embedding Model Configuration

• EMBEDDING_PROVIDER (Optional): Choose 'ollama' (default) or 'openai'.
• EMBEDDING_MODEL (Optional):
  • When using Ollama, the default is 'nomic-embed-text'.
  • When using OpenAI, the default is 'text-embedding-3-small'.
• OLLAMA_URL (Optional): The URL address of the Ollama instance (default is http://localhost:11434).
• OPENAI_API_KEY (Required if using OpenAI): Your OpenAI API key.

Development Guide

Clone the repository

git clone https://github.com/qpd-v/mcp-server-ragdocs.git
cd mcp-server-ragdocs

Install dependencies

npm install

Build the project

npm run build

Run locally

npm start

Troubleshooting

Common Issues

1. Qdrant Connection Error

   Error: Unable to connect to the Qdrant service at http://localhost:6333
   

   • Check if Docker is running.
   • Confirm if the Qdrant container is running properly: docker ps | grep qdrant.
   • Try restarting the container.

2. Ollama Model Not Found

   Error: The model nomic-embed-text was not found.
   

   • Run: ollama pull nomic-embed-text.
   • Verify if the model is installed: ollama list.

3. Configuration Path Issue

   • On Windows systems, ensure the correct path separator (backslash) is used.
   • Check if the variables in the .env file are set correctly.

4. Port Conflict

   Error: EADDRINUSE @localhost:5000
   

   • Check if another program is using port 5000.
   • Try changing to another available port.

Contribution Guide

Submit Code

1. Create a feature branch:

   git checkout -b feature/your-feature-name
   

2. Commit changes:

   git add .
   git commit -m "Describe the modifications you've made"
   git push origin feature/your-feature-name
   

3. Create a Pull Request to the main branch.

Report Issues

• Before submitting an ISSUE, search for existing issues to avoid duplicates.
• Provide detailed error information and steps to reproduce the problem.

💻 Usage Examples

Basic Usage

In Claude Desktop or other MCP-compatible clients:

1. Add documentation:

Add this documentation: https://docs.example.com/api

2. Search for documentation:

Search for documentation to get information about authentication

3. List sources:

What are the available documentation sources?

📄 License

This project is licensed under the MIT License.

💡 Usage Tip

For more information, please visit our GitHub repository.