Top tracks and top artists over a period sit in the same conversation as playlist creation, so "build me a playlist of what I actually played this month" is one exchange rather than an afternoon. It is something you run yourself, on your own Spotify developer credentials, on your own machine.
A Python server built on Gradio that reaches Spotify through a developer app of your own. It runs as a local Gradio application and clients connect to its SSE endpoint rather than launching a process, so the app has to be up before the client can see anything. The listening-history tools are the part that distinguishes it: top artists and top tracks over a period, in the same conversation that creates playlists.
- OAuth sign-in to your own Spotify account
- Track search returning the artist alongside the track
- A song added to your playback queue
- Your recently played tracks
- A new playlist created, your existing playlists listed with their ids, and songs added to any of them
- Your top artists and top tracks for a period, with genre information
Python, and a Spotify developer app of your own with a loopback redirect URI registered on it — the client ID and secret go in a local .env file. There is no published package: clone the repository, install its two dependencies and run it, then point the client at the Gradio SSE endpoint through mcp-remote. Queueing a song needs Spotify Premium; search, playlists and listening history do not. The project notes that the client may need restarting once or twice before the connection takes, and that it is built to run locally rather than in an ephemeral environment.