This project provides a YouTube search application designed to demonstrate three distinct methods for accessing and searching YouTube content. It serves as an educational resource and reference implementation, showcasing direct Flask app integration with the YouTube Data API, a dedicated MCP server for structured search, and a Flask client that consumes the MCP server. While not intended for production use, it is an excellent tool for developers to understand MCP concepts, explore different architectural patterns for API interaction, and integrate YouTube search capabilities into AI clients like Claude Code or Gemini CLI.