youtubeinsights

The YouTube Insights MCP Server exposes tools that let agents extract transcripts, search for videos by keyword with metadata, and retrieve channel information from YouTube URLs. It supports multi-language subtitles and provides fast integration via the FastMCP-based tooling. To use it, deploy the server with the uvx workflow described in the installation section, then invoke the MCP tools such as get_youtube_transcript, search_youtube_videos, and get_channel_info through your MCP client or agent workflow. The transcript tool returns subtitles (e.g., Korean and English) for given video URLs, the search tool returns video metadata including views, likes, and thumbnails, and the channel info tool fetches the channel’s metadata and recent uploads based on a video URL.

Prerequisites:

• A YouTube Data API key with access to video transcripts and metadata
• Node.js or Python-enabled environment as applicable to your deployment (this server uses uvx/uv for Python-based deployment)
• Access to install dependencies from your automation environment

Installation steps (UVX/Claude-friendly workflow):

1. Prepare your API key and environment
   • Obtain a YouTube Data API key and store it securely.
2. Install via uvx or set up development environment

   • Using uvx (recommended):

     • Ensure uvx is installed: Follow Astral UV docs to install uvx.

     • Deploy configuration (example shown in the repo's docs): Add to your MCP settings file (e.g., claude_desktop_config.json):

       { "mcpServers": { "youtubeinsights": { "command": "uvx", "args": ["youtubeinsights-mcp-server"], "env": { "YOUTUBE_API_KEY": "your-api-key" } } } }

   • Alternative development run (local):

     • Copy the repository and fill in credentials in a .env file.
     • Run the server with the UV tool: uv --directory path/to/youtubeinsights-mcp-server run youtubeinsights-mcp-server Ensure the environment variable YOUTUBE_API_KEY is set in your shell or .env.
3. Verify the server is running by invoking a test MCP tool command via your agent or CLI.

Note: If you prefer using npx/Smithery distribution, you can install via npm as shown in the README, then configure your MCP client accordingly.

Environment variables and configuration:

• YOUTUBE_API_KEY is required for API access. Keep it secure and never commit it.
• For uvx-based deployments, you can pass the key via the env map in your MCP settings file.
• The server supports multiple languages for transcripts; ensure your usage locale aligns with your needs (e.g., en, ko).

Common issues:

• 403 or quota errors from YouTube API: verify API key permissions and quota limits.
• Network/firewall restrictions preventing outbound API calls: allow requests to YouTube APIs.
• Misconfigured command/args in MCP settings: ensure you reference the correct executable and server name (youtubeinsights-mcp-server).

Troubleshooting tips:

• Check logs for gaps between starting the server and tool availability.
• Validate environment variables are loaded in the runtime environment.
• Use sample tool descriptions to confirm the expected JSON inputs/outputs for each MCP tool.