This MCP server exposes the YouTube Free Deep Research CLI as a programmable MCP service for AI assistants. It integrates with Claude Desktop workflows and related tooling to enable deep research workflows, multi-source content processing, and podcast-style content generation via the YouTube data pipeline. Use the MCP entrypoint to connect to the server from your agent/client, then call the available endpoints or workflows exposed by the accompanying FastAPI backend to manage YouTube data retrieval, transcript processing, and TTS-enabled output generation. The server is designed to work in tandem with the Next.js dashboard and the modular research stack, providing streamlined access to RAG, content extraction, and multi-engine TTS capabilities through a consistent MCP interface.

Prerequisites:

• Node.js and npm installed on the host (required for npx to run MCP server).
• Optional: Docker if you prefer containerized usage.

Install and run:

1. Ensure Node.js and npm are available: node -v npm -v

2. Run the MCP server using npx (no local installation required): npx -y jaegis-youtube-chat-mcp

3. (Optional) Run via Docker: docker build -t jaegis-youtube-chat-mcp . docker run --rm -p 3000:3000 jaegis-youtube-chat-mcp

4. If you are deploying in a local dev environment and need the API server, ensure dependencies are installed for the backend as described in the repository (Python/uvicorn steps are referenced in the project docs if you also run the Python API alongside):

   For Python API backend (not required for the MCP runtime itself)

   python -m venv venv source venv/bin/activate pip install -r requirements.txt uvicorn youtube_chat_cli_main.api_server:app --reload --port 8556

Tips and caveats:

• The MCP server relies on the jaegis-youtube-chat-mcp package for the MCP interface. Ensure you are using a compatible Node.js version and that your environment can run npx without network restrictions.
• If you rotate OpenRouter keys or other API credentials, store them securely and pass them to the server through environment variables as needed by the backend services (for example, OPENROUTER_API_KEYS for rotation).
• For production deployments, consider using the Docker image or a process manager to keep the MCP server running (e.g., systemd, PM2).
• If you encounter issues with YouTube data access limits, adjust rate limiting settings in your workflow and utilize the built-in intelligent queuing features of the underlying system.
• The MCP server supports integration with Claude Desktop workflows; ensure your client is configured to call the correct MCP endpoint and that authentication/authorization flows align with your environment.