Digarr reached 300 stars on GitHub on October 7, the day after I wrote about v1.19.0.
I started it because I wanted suggestions for new artists. I wasn’t using Spotify or the other commercial streaming platforms, but I still wanted help finding music based on what I already liked.
So I built Digarr. It’s a self-hosted music discovery app: connect your listening history or library, get suggestions, listen to previews, and decide what to keep. You can send your picks to Lidarr, slskd, or playlists. You bring your own AI provider or local model.
I’ve worked in IT for over 20 years, and bringing agentic AI into my workflow helped tremendously. What I’d expected to take about three years took three months.
And people use it and like it. It turns out I wasn’t the only one missing this. There have been 211 issues and 580 pull requests in total, and the Docker Hub image has been pulled over 73,000 times.
I still use Digarr almost every day. I’ll keep improving it, fixing things, and maintaining it. I’ve also overhauled the documentation, from the README to the deployment and recovery guides. The instructions match v1.19.0 more closely now, with more detail on backups, credentials, and known limitations.
The response to Digarr gave me the confidence to start other projects and share them. Garret is an offline writing studio for novelists. Ditero is a self-hosted task app for families, friends, and small groups. Both are free and open source on GitHub.
If you’ve tried Digarr or sent feedback, thank you. The feedback helps me fix things, and it’s nice to know something I made is useful to someone else.
And if it helped you find an artist you now listen to, I’d like to hear about that. That was the reason I started.