MCP-youtube

This MCP server exposes a comprehensive set of YouTube analysis and AI-powered text tools via MCP endpoints and REST API. You can search YouTube videos, analyze comments, generate scripts, extract SEO insights, compare videos, and perform a suite of AI text operations (rewrite, summarize, expand, translate, and more). The server is designed to be called from web browsers (CORS enabled) and supports MCP protocol calls as well as direct REST requests. Typical usage involves calling the MCP endpoint to list tools or to invoke a specific tool by name with the required arguments, enabling seamless chaining of data gathering, processing, and content generation tasks for YouTube-focused workflows.

Prerequisites:

• Node.js (recommended latest LTS) or access to a Cloudflare Worker deployment workflow
• npm or yarn
• Optional: Cloudflare Wrangler CLI if deploying as a Worker (wrangler login, wrangler config)

1. Install Wrangler (for Cloudflare Workers) and login

npm install -g wrangler
wrangler login

2. Retrieve the MCP YouTube server package

• If you are using the provided npm package or repository, follow the repository's install steps. The MCP server is exposed at a Cloudflare Worker URL, so you may deploy via Wrangler or a similar workflow depending on your environment.

3. Configure deployment

• For Wrangler-based deployments, ensure wrangler.toml points to the correct account, zone, and worker script. If this MCP server ships as a Worker, you may publish it directly:

wrangler publish

4. Run locally (if supported by the project)

• Some projects provide a local dev server mimic. Check repository for server.js or index.js entry. If available, start with:

node server.js

5. Verify deployment

• After starting, test the MCP endpoint at the provided URL (for example, https://your-worker.your-namespace.workers.dev/mcp).
• Also test REST endpoints like /youtube/search and /mcp to confirm both REST and MCP protocol work as expected.

Notes and tips:

• The MCP endpoint supports the standard tools/list and tools/call methods. Use tools/list to discover available tools and their exact parameter names.
• CORS is enabled, so you can call the REST endpoints directly from browser-based apps.
• The server implements a KV-based caching layer to accelerate repeated queries; configure cache TTL based on your usage patterns.
• If deploying to Cloudflare Workers, ensure your environment variables (if any) are set in wrangler.toml or your host platform. Monitor quota limits for R2 KV or similar if used.
• The npm_package field is null because this deployment is described as a Worker-based MCP server rather than a traditional Node.js package installable via npm.