Work can continue while you are away
Workers use the running cloud computer. Scheduled scripts can carry out recurring jobs, but they still need valid connections, available resources and provider allowance. A website verification or other human step can interrupt a job.
For a recurring task, ask the worker to save its schedule, output location and failure handling in the project. Check an initial run before relying on the schedule.
Recover saved work
Wrkr backs up worker homes and saved local state off-server, with cache exclusions. Ask your worker to recover a lost file or an earlier version. Retrieved files appear as separate, dated copies so you can inspect them without overwriting current work.
Recovery requires a retained backup containing the file. Normal backups run every 30 minutes. Heavy changes above the plan’s backup allowance reduce copying to daily and retain the latest two copies. The account page shows backup age, history and recovery information.
History and earlier-file downloads have plan-dependent limits. Machine-loss recovery is separate from the file-download allowance. Backups do not promise uninterrupted service or permanent storage.
Download important files
Use Files to download work you need to keep independently. Check that the download opens and contains what you expected before removing a worker.
Post-cancellation export and deletion deadlines are being finalized before paid launch. Ask support about data requests. A backup is not a substitute for your own copy of important work.
What is still being prepared?
Paid billing, cancellation controls and post-service data handling are launch work. Native Android and iOS apps are planned; current access is through the browser. Detailed app-service setup guides will follow verification of those services.