Monitor a source
Search for available content or create a watch for a future match.
System overview
How does autofetch work?
autofetch connects centrally managed watches with downloads on your own device. The portal finds matching content and organises jobs; a connected client downloads files directly to your computer, NAS or Raspberry Pi.
A shared workflow
From source to local file
Find a match
autofetch checks the selected source and creates a job when the rules match.
Download locally
Your client receives the job and downloads directly from the source.
Report the result
Progress, success and understandable errors remain visible in the portal.
More than a download link
From one-off search to a persistent watch
A manual search starts with content that is already available. A watch can start earlier: define what you are looking for and let autofetch check again later.
autofetch distinguishes not available yet from technically failed. Future content is not treated as a broken download.
Series and numbered releases
Follow ongoing content automatically
Supported sources can insert a counter into a search pattern. A watch such as Example S02E{counter} can search for Example S02E01, Example S02E02 and later episodes in sequence.
You can set the start value, step, formatting and an optional end value. If episode 5 appears before episode 4, the later episode may be detected while the gap remains open.
Central coordination
What the portal does
- Manage searches and watches
- Check configured sources
- Turn matches into download jobs
- Assign jobs to connected clients
- Display progress, results and errors
Local execution
What your client does
- Receive jobs securely
- Download directly from the selected source
- Store files on your device
- Report progress and results
- Resume supported HTTP downloads after interruptions
The actual files travel directly from the source to your device.
One model for different sources
Different technologies, one understandable process
Watch → match → job → local client → result
Media libraries, feeds, release pages and IRC/XDCC work differently, but search, retries, progress and error handling stay together.
What autofetch does not provide
- No media catalogue and no bundled content sources
- No public self-hosting documentation for the server
- No central storage or relay of downloaded media
- No guarantee that an external source allows or continues to provide a download
You configure and use sources for which you have the required permission.
Open client, centrally operated server
The client is published separately on GitHub. The autofetch server is currently operated centrally and is not offered as a public self-hosting product.