home / mcp / github mcp server

GitHub MCP Server

a mcp server to manage github project's functionality

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "kunwarvivek-mcp-github-project-manager": {
      "command": "mcp-github-project-manager",
      "args": [],
      "env": {
        "GITHUB_REPO": "<your_repo>",
        "GITHUB_OWNER": "<your_owner>",
        "GITHUB_TOKEN": "<your_github_token>",
        "GOOGLE_API_KEY": "<google_api_key>",
        "OPENAI_API_KEY": "<openai_key>",
        "ANTHROPIC_API_KEY": "<anthropic_key>",
        "PERPLEXITY_API_KEY": "<perplexity_key>"
      }
    }
  }
}

You can run the MCP GitHub Project Manager to transform project ideas into actionable GitHub tasks with end-to-end traceability. This server integrates with GitHub’s Projects (v2) and AI-assisted tooling to generate PRDs, break down work, and track changes across requirements, features, use cases, and tasks.

How to use

You interact with the MCP server through a client that speaks the Model Context Protocol. Your workflow typically starts with creating a project, generating a PRD from your idea, parsing that PRD into concrete tasks, and then managing those tasks with traceability links from requirements to implementation. Use the client to invoke tools like PRD generation, task parsing, feature addition, and traceability creation. The server provides real-time sync with GitHub and rich context to guide developers from business goals to code changes.

How to install

# Prerequisites
- Node.js (LTS version) and npm installed
- Docker installed (optional but recommended for containerized runs)

# Option A: Install the MCP server globally via npm
npm install -g mcp-github-project-manager

# Start the MCP server locally (stdio transport)
mcp-github-project-manager

# If you prefer to run with environment variables in one command
GITHUB_TOKEN=your_github_token GITHUB_OWNER=your_owner GITHUB_REPO=your_repo mcp-github-project-manager

# Option B: Run with Docker
# Build the image from a Dockerfile if you maintain a custom image, or pull an official image if available
# Example run with required environment variables
docker run -it \
  -e GITHUB_TOKEN=your_github_token \
  -e GITHUB_OWNER=your_owner \
  -e GITHUB_REPO=your_repo \
  mcp-github-project-manager

Configuration and prerequisites

Configure environment variables to enable GitHub access and AI capabilities. You must provide a GitHub token with repository, project access, and organization permissions if you plan to manage projects and issues. At least one AI provider key is required for AI-powered features. Additional context and traceability settings can be tuned to your needs.

Configuration examples

# Example environment file (.env)
GITHUB_TOKEN=ghp_exampleToken1234567890
GITHUB_OWNER=your-github-username
GITHUB_REPO=your-repo-name

# AI providers (at least one required for AI features)
ANTHROPIC_API_KEY=sk-example-anthropic-key
OPENAI_API_KEY=sk-example-openai-key
GOOGLE_API_KEY=your-google-api-key
PERPLEXITY_API_KEY=your-perplexity-key

# Optional task/context settings
ENHANCED_TASK_GENERATION=true
AUTO_CREATE_TRACEABILITY=true
ENHANCED_CONTEXT_LEVEL=standard

Troubleshooting

If you encounter issues starting the MCP server, verify that your environment variables are correctly set, the GitHub token has the required permissions, and you are using compatible Node.js and Docker versions. Check that the server process starts without binding errors, and ensure the environment is able to reach GitHub and any configured AI provider endpoints.

Notes

The MCP server supports stdio and HTTP transports. Use stdio for local development and HTTP if you deploy a remote MCP endpoint. Real-time synchronization with GitHub is supported, and the system includes end-to-end traceability for requirements, features, use cases, and tasks.

Available tools

generate_prd

Transform project ideas into comprehensive PRDs using AI providers.

parse_prd

AI-powered parsing of PRDs into actionable development tasks with traceability.

add_feature

Add new features with automatic impact analysis and task generation.

analyze_task_complexity

AI-based analysis of task complexity, effort estimation, and risk.

get_next_task

AI-powered recommendations for optimal task prioritization.

expand_task

Break down complex tasks into subtasks automatically.

enhance_prd

AI-driven PRD enhancement with gap analysis.

create_traceability_matrix

End-to-end traceability from PRD requirements to tasks with bidirectional links.