GitHub MCP Server

A FastMCP server that provides access to GitHub repositories and data for Poke integration.

🚀 Features

• get_repos: Get repositories for a GitHub user
• get_issues: Get issues for a repository
• get_pull_requests: Get pull requests for a repository
• search_code: Search for code on GitHub

🔑 GitHub Token Setup

This server requires a GitHub personal access token for API access:

1. Go to GitHub Settings > Personal Access Tokens
2. Generate a new token with repo and public_repo scopes
3. Set the environment variable: export GITHUB_TOKEN=your_token_here

🛠️ Local Development

# Install dependencies
pip install -r requirements.txt

# Set GitHub token
export GITHUB_TOKEN=your_token_here

# Run the server
python src/server.py

🚢 Deployment

Option 1: One-Click Deploy to Render

Option 2: Manual Deployment

1. Fork this repository
2. Connect your GitHub account to Render
3. Create a new Web Service on Render
4. Connect your forked repository
5. Set environment variable GITHUB_TOKEN
6. Deploy!

Your server will be available at https://github-mcp.onrender.com/mcp

🎯 Poke Integration

1. Go to poke.com/settings/connections
2. Add the MCP URL: https://github-mcp.onrender.com/mcp
3. Give it a name like "GitHub"
4. Test with: "Tell the subagent to use the GitHub integration's get_repos tool"

🔧 Available Tools

• get_repos(username, limit=10): Get repositories for a user
• get_issues(owner, repo, state="open", limit=10): Get issues for a repository
• get_pull_requests(owner, repo, state="open", limit=10): Get PRs for a repository
• search_code(query, language="", limit=10): Search for code

📝 Example Usage

# Get user's repositories
get_repos(username="octocat", limit=5)

# Get open issues
get_issues(owner="microsoft", repo="vscode", state="open", limit=10)

# Get pull requests
get_pull_requests(owner="facebook", repo="react", state="open", limit=5)

# Search for Python code
search_code(query="machine learning", language="python", limit=10)

⚠️ Rate Limits

GitHub API has rate limits:

• With token: 5,000 requests per hour
• Without token: 60 requests per hour

The server will return an error if rate limits are exceeded.