YouTube MCP Server

This project connects YouTube video processing capabilities to Claude AI via the Model Context Protocol. It uses yt-dlp for downloading subtitles and ffmpeg for screenshot extraction. Users can ask Claude to summarize YouTube videos by providing the URL.

System Architecture Diagram

graph TD
    A[User] --> B[Claude AI]
    B --> D[mcp-youtube]
    D --> E[yt-dlp Subtitles]
    D --> F[ffmpeg Screenshots]
    D --> G[Temporary Storage]
    G --> H[Auto-Cleanup]

    style A fill:#FFE4B5,stroke:#333
    style B fill:#98FB98,stroke:#333
    style C fill:#ADD8E6,stroke:#333
    style D fill:#D0ECE7,stroke:#333
    style E fill:#FFB6C1,stroke:#333
    style F fill:#DDA0DD,stroke:#333
    style G fill:#FFFACD,stroke:#333
    style H fill:#FFD700,stroke:#333

    subgraph MCP
        D --> E
        D --> F
        D --> G
    end

    subgraph Temporary Storage
        G --> H
    end

Getting Started

Prerequisites

• yt-dlp and ffmpeg installed (via Homebrew, WinGet, or other package managers)
• Node.js and npm installed

Installation

1. Install the MCP server:

   npm install -g @novemberde/mcp-youtube

2. Configure Claude to use the MCP server: Add the following to your claude_desktop_config.json:

   {
     "mcpServers": {
       "mcp-youtube": {
         "command": "npx",
         "args": ["@novemberde/mcp-youtube"]
       }
     }
   }

Available Tools

1. download_youtube_url

Downloads and extracts subtitles from a YouTube video.

Parameters:

• url (required): YouTube video URL

Example:

{
  "url": "https://www.youtube.com/watch?v=example"
}

2. search_youtube_videos

Searches YouTube videos based on a query.

Parameters:

• query (required): Search term
• max_results (optional, default=10): Maximum number of results

Example:

{
  "query": "how to make pizza",
  "max_results": 5
}

Returns:

• Video title, URL, description, duration, view count, uploader name

3. get_screenshots

Captures screenshots from specific timestamps of a YouTube video.

Parameters:

• url (required): YouTube video URL
• timestamps (required): Array of timestamps in HH:MM:SS format

Example:

{
  "url": "https://www.youtube.com/watch?v=example",
  "timestamps": ["00:01:30", "00:05:45"]
}

Returns:

• File paths to saved screenshots

Notes

• Ensure yt-dlp and ffmpeg are installed globally.
• The server uses temporary directories for processing, which are automatically cleaned up.
• Error handling provides descriptive messages for troubleshooting.

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/your-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin feature/your-feature)
5. Create a new Pull Request

License

This project is licensed under the MIT License - see the COPYING file for details.