How It Works
reklawdbox sits between your AI agent and Rekordbox. It reads your database, caches enrichment and audio analysis data, and stages Rekordbox metadata for XML export. Your database is never modified directly. Separate file tools can write tags or artwork to audio files when called.
The basics
Section titled “The basics”- The agent reads your library — track metadata, playlists, play history. All read-only.
- Enrichment and audio analysis get cached — provider metadata and audio features persist in a local SQLite store outside Rekordbox.
- Rekordbox metadata is staged in memory —
update_trackschanges can be previewed or discarded before XML export. - Direct file tools stay explicit — tag and artwork operations can write audio or image files without changing
master.db. - You export and import XML — export writes a local XML file and runs the backup step; you choose whether to import it into Rekordbox.
That’s the whole loop. The agent handles the tedious work, you make the decisions.
Caching
Section titled “Caching”Most workflows need cached data to work well. The Library Cleanup guide walks you through populating caches as part of the process. Fresh cache entries are reused instead of repeating enrichment or analysis.
Provider enrichment and audio-analysis cache rows are reconstructible, but the same internal SQLite store also holds durable audit resolutions and notes, saved presets, calibration/profile statistics, and broker session metadata. Treat it as persistent application data: moving or deleting it loses those decisions and settings. Setup can write MCP host and broker configuration, and backups and XML exports create additional local files.