A self-hosted catalogue that identifies every album in your archive down to the pressing and corrects tags without breaking anything.
bash -c "$(curl -fsSL https://raw.githubusercontent.com/xooxoxxo/tagave/main/install-tagave.sh)"
Needs Docker with Compose 2.20 or later. Free and open source, AGPL-3.0.
For a collection you actually own: folders ripped over twenty years, tags from a dozen programs, and a shelf of records next to them.
Every album is matched against MusicBrainz and Discogs, with acoustic fingerprinting when the tags are wrong or missing. It finds the edition you have, with its catalogue number, and knows that folders named CD1 and CD2 are one release.
Changes are proposed as a plan and previewed file by file before anything is written. Every write is journalled and can be reverted. A field whose correct value is unknown is never blanked, and files are never moved, split or converted.
Incomplete albums, duplicates and missing releases by artists you follow are listed for you. Sync a Discogs collection and the vinyl and CDs on your shelf join the same catalogue as the files.
A plan lists each file, each field, the value it has now and the value it would get, and why. You apply it when it looks right, or change the policy and preview again.
A compilation filed one folder per artist turns into a row of one-track albums. tagave notices the pieces that share a title and fit together, treats them as one album, and sets the album values for every track in one previewed plan.
Once albums are identified, tagave can tell which ones are short of tracks, which are duplicated across formats or editions, and which releases by artists you follow are not in the library at all.
The installer pulls the published images, asks where your music is, generates the secrets and starts everything in ~/tagave. Running it again is safe.
bash -c "$(curl -fsSL https://raw.githubusercontent.com/xooxoxxo/tagave/main/install-tagave.sh)"
Requires Docker and Docker Compose 2.20 or later, on Linux or macOS.
Answer where your music lives. Every question also has a flag, so it can run unattended.
Go to http://localhost:3100 and create your account.
When the setup asks for a scan root, enter /mnt/music. That is your music folder as the file worker sees it.
tagave catalogues one real library every day. It has one user so far, so expect rough edges, and read the operations guide before you trust it with anything you cannot re-rip.
A catalogue for a music collection you own. It scans your folders, identifies each album against MusicBrainz and Discogs, shows what is incomplete or duplicated, and corrects tags through plans you preview first.
No. Navidrome, Jellyfin, Plexamp and Roon play your music, and they work better when the tags underneath are right. tagave is the part that gets them right.
Only their tags, only when you apply a plan, and only after you have seen the preview. Each write is journalled so it can be reverted, and the audio is hashed before and after so a write that touched the audio stops the job.
Files are never moved, renamed, split, converted or deleted. You can also mount your music read-only and use tagave as a catalogue alone.
No. It runs on your own computer or server in Docker, and your library and account stay there. It talks to the metadata services it uses (MusicBrainz, Cover Art Archive, Wikipedia and Wikidata, plus Discogs and AcoustID if you add keys for them). If you set a release feed under Settings › Updates, it checks that feed for new versions when you open that page or press Check now. Nothing else leaves your server, and there is no telemetry.
It is free. The code is open source under the GNU AGPL-3.0. The tag writer runs mutagen as a separate process.
It reads and writes tags in MP3, FLAC, M4A and ALAC, Ogg, Opus, WAV and AIFF. APE and WavPack tags are read but not written yet, and albums that come from a CUE sheet only get album-level fields.