Interest Project
Movie Recommendation Project
This project combines my interest in film with AI-assisted tooling. It uses a Python CLI agent connected to a local Letterboxd MCP server so the system can read Letterboxd context and generate movie recommendations through Gemini.
What makes it interesting to me
- It connects a real personal interest, film watching and logging, with agent-based system design.
- It bridges Python and Node.js through MCP with HTTP/SSE transport.
- It turns taste, history, and mood into a recommendation workflow rather than a static watchlist.
Current implementation
- Python CLI agent for recommendation flow.
- Letterboxd MCP server integration with protected write actions.
- Environment-based setup for Gemini, TMDB, and Letterboxd authentication.