mcp-github-project-manager

The MCP GitHub Project Manager server provides AI-powered task management and complete traceability for GitHub projects. It can generate PRDs, parse PRDs into tasks, add features with impact analysis, analyze task complexity, and recommend next tasks. It also supports end-to-end traceability from business requirements to features, use cases, and tasks, with configurable context depth and optional AI-powered enhancements. To use it, start the MCP server (via npx as shown in the config) and supply the required GitHub context through environment variables. Once running, you can interact with the server through its MCP endpoints to manage GitHub Projects, Issues, Milestones, and related tasks, and you can enable AI providers to access advanced reasoning and generation features for planning and execution.

Prerequisites:

• Node.js and npm installed on your system
• Basic knowledge of environment variable configuration

Step 1: Install the package via MCP command (or globally as needed)

• If using MCP config, the server can be started by running the command in your MCP runner:

# Install and run via npx (shown in mcp_config)
# This will fetch and run the package as part of MCP workflow

Step 2: Set up environment variables

• Create and configure an environment file or export variables in your shell:

export GITHUB_TOKEN=your_github_token
export GITHUB_OWNER=your_github_username_or_organization
export GITHUB_REPO=your_repository_name

# Optional AI providers
export ANTHROPIC_API_KEY=your_anthropic_api_key_here
export OPENAI_API_KEY=your_openai_api_key_here
export GOOGLE_API_KEY=your_google_api_key_here
export PERPLEXITY_API_KEY=your_perplexity_api_key_here

Step 3: Run the MCP server

• If using the provided MCP config, the runner will start the server as configured.
• Alternatively, install locally and start the CLI as directed by your setup.

Step 4: Validate access

• Open your MCP client or API endpoint to verify the server is responding and that GitHub context is accessible.

Notes:

• Ensure the GitHub token has repo, project, and write:org scopes for full functionality.
• If AI keys are unavailable, the server will gracefully fall back to traceability-based context.
• The server supports multiple AI providers; configure at least one to enable full AI features.
• For Docker or containerized deployments, ensure environment variables are passed to the container instance.
• Review rate limits on GitHub GraphQL API usage and implement appropriate backoff strategies.
• Use the traceability features to identify orphaned tasks and assess coverage between PRDs, features, and tasks.