About
How Music Mirror works
A non-commercial project built on open music data. Your listening history stays in your browser.
Privacy
- Your history never leaves your browser. The ZIP is unzipped and parsed client-side; the last-24h API slice is fetched with your token directly from Spotify. Your plays are never uploaded to our servers.
- Song matching.To match songs, we send track and artist names (never your listening times or play counts) through our server to MusicBrainz, ListenBrainz and ReccoBeats. We don't store or log them.
- No raw plays are stored. The grid is a derivative — one colour per day, computed from your local session. Close the tab and the plays are gone.
- What you see is what the mint will capture. The image embedded in your pass is the grid above; no per-track list, no timestamps, no user ID.
- User-delete-on-demand is the default. Clear the page with the clear button at the top; hard-reset the browser token at
/mirrorto sever the Spotify link.
Data sources & credits
- MusicBrainz core data · CC0 1.0
Recording/artist MBIDs, first-release dates, credits, relationships, search - MusicBrainz tags and genres · CC BY-NC-SA 3.0
Artist/recording genres (on by default for the non-commercial launch) - ListenBrainz API · CC0 1.0
Bulk recording/artist metadata, public profile listens - ListenBrainz Labs API · CC0 1.0 (MetaBrainz experimental service)
Recording search, similar artists, MBID to Spotify-ID mapping - Wikidata · CC0 1.0
Artist genres (P136 via MusicBrainz artist ID P434) - ReccoBeats API · ReccoBeats terms of service
Audio features (key, mode, tempo, energy, valence, acousticness) - Relationships seed (data/relationships.seed.json) · Facts with a cited public source per entry
A handful of hand-curated sample/feature relations