GitHub MCP Server Claude

Overview

GitHub MCP Server is an API-based tool that interacts with GitHub using the MCP (Modular Command Processor) framework. It provides various functionalities, including fetching user details, repository information, and authenticated user data using GitHub's REST API.

This project is built using Python and leverages httpx for asynchronous API calls. It also uses dotenv for secure handling of GitHub authentication tokens.

Features

• Fetch GitHub user information by username.
• Retrieve details of a GitHub repository.
• Get authenticated user details using a GitHub personal access token.
• Utilizes FastMCP for modular command processing.

Project Structure

D:/MCP_Project/ 
|----.env
│----.gitignore
│----claude_desktop_config.json  (Create this file in C:\Users\your_username\AppData\Roaming\Claude\)
│----main.py
│----pyproject.toml
│----README.md
│----requirements.txt

• main.py: Core logic of the GitHub MCP Server.
• .env: Stores environment variables (e.g., GitHub Token).
• claude_desktop_config.json: Configuration for running the MCP Server.
• requirements.txt: Lists required dependencies.
• explanation_video.mp4: A video explaining the project.

Setup Instructions

Prerequisites

• Python >=3.10
• GitHub personal access token (for authenticated requests)
• conda or venv for virtual environment management

Setup and Usage

1. Clone the Repository

   git clone https://github.com/DivyanshKushwaha/GitHub-MCP-Server-Claude.git
   cd GitHub-MCP-Server-Claude
   

2. Create Python environment

python -m venv venv
source venv/bin/activate  

3. Install Dependencies

pip install -r requirements.txt

4. Set Up Environment Variables (Create .env file)

GITHUB_TOKEN=your_personal_access_token

5. Setup claude_desktop_config.json

{
    "mcpServers": {
        "MCP_Server": {
            "command": "my_env/Scripts/uv",
            "args": [
                "run",
                "D:/MCP_Project/main.py"
            ]
        }
    }
}

• The command key specifies the path to the uv script located in the conda environment. This is used to run the server.

• The args key provides additional arguments for the uv script:

  • "run": Indicates the action to run the server.
  • my_env : python environment 'my_env'
  • "D:/MCP_Project/main.py": Specifies the path to the main.py script, which contains the implementation of the MCP server.

6. Launch the Claude Desktop Application

• Open the Claude Desktop Application.
• It will use the MCP Server as configured in the claude_desktop_config.json file to fetch and process data.