Podcast RSS ingestion

The user-facing part.

Turn podcast feeds into a searchable text library. You add RSS URLs in your settings panel, and the helper polls for new episodes in the background. It downloads the audio files, runs local Whisper transcription, and applies speaker diarization. The final output is formatted as a standard markdown corpus, grouped by show and stored in your library.

Run the loop.

1. Open the Uoink desktop settings.
2. Paste a podcast RSS feed URL into the field.
3. Click Add to register the feed.
4. Select your preferred Whisper model size.
5. Let the helper download and transcribe the latest episodes in the background.

How Uoink handles it.

The helper polls registered feeds periodically using a feedparser client. It checks new episodes against the SQLite index to prevent duplicate downloads. The audio is downloaded locally, and Whisper performs the transcription. If WhisperX is configured, it runs a diarization pass to label speakers before writing the markdown file to your disk.

Agent-readable surface.

Full server metadata lives at /mcp and /mcp/manifest.json.

Run it on your next source.

Add your favorite podcast RSS feeds to build a local audio transcript archive.

Install Uoink