File checks

A file check validates an artifact without importing it into a database. Use it for non-database backups or as a lightweight freshness and transfer check.

What a file check can verify

• a matching file exists
• the selected file is no older than the configured maximum age
• its size is at or above a minimum and/or at or below a maximum
• the project worker can download it
• an age-encrypted file can be decrypted with the selected identity or passphrase

Create a file schedule

1. 1

   Choose the worker and source

   Select a project worker with network access to the source and enough scratch space for the downloaded and decoded file.

2. 2

   Select the file

   Choose automatic latest-file selection or provide a filename pattern with UTC date variables. Use / in source-relative paths. If a pattern can match several files, optionally choose the latest by modification time.

3. 3

   Add freshness and size bounds

   Enable maximum age to reject stale files. Add a nonzero minimum size to catch empty or unexpectedly small output, and a maximum when unusually large artifacts should fail.

4. 4

   Configure age decryption if required

   Select age identity or age passphrase and its secret metadata. File checks do not use ZIP-password handling because they do not open archive containers.

5. 5

   Set cadence and notifications

   Choose an allowed cadence and UTC time, then select how success or error notifications should be delivered.

Run the schedule manually with a known-good artifact, then test that your criteria detect a deliberately stale or out-of-range example where practical. This confirms the thresholds express your intended backup policy.

See Checks for statuses and Evidence for what a completed file check can establish.