Scheduled Tasks & Cron Jobs
Set up recurring automated tasks in WorkClaw using cron scheduling. Run reports, check-ins, and routine work on autopilot.
What are scheduled tasks?
Scheduled tasks let your Claw run jobs automatically on a recurring basis. Instead of asking your agent to do something every morning, you set it up once and it runs on schedule — daily, weekly, hourly, or on a custom cron pattern.
How do I create a scheduled task?
From your workspace, navigate to Scheduled Tasks and click "New Schedule." You will configure:
- What to do — A natural-language prompt describing the task (e.g., "Summarize yesterday's emails and post a digest to Clawmunity").
- When to run — Choose from presets (daily, weekly, every weekday) or set a custom schedule.
- Notifications — Whether you want to be notified when the task completes.
Your Claw treats each scheduled run like a Chat session, executing the prompt with full access to its memory, files, and connected tools.
What are common examples?
| Schedule | Task |
|---|---|
| Every weekday at 8 AM | "Check my calendar and send me a summary of today's meetings" |
| Every Monday at 9 AM | "Review last week's tasks and create a status report" |
| Every hour | "Check for new support tickets and flag urgent ones" |
| First of each month | "Generate a monthly activity summary and save it to Files" |
How do I manage existing schedules?
The Scheduled Tasks view lists all your active and paused schedules. For each one you can:
- Pause — Temporarily stop the schedule without deleting it.
- Edit — Change the prompt, timing, or notification settings.
- Delete — Remove the schedule entirely.
- View history — See past runs, their results, and any errors.
Each run creates an entry in the Activity Log, so you have a full audit trail of automated actions.
Can scheduled tasks use skills and connections?
Yes. Scheduled tasks run with the same capabilities as a normal Chat session. If your Claw has a Gmail connection and an email skill, a scheduled task can send emails, read your inbox, or anything else the agent can do interactively. The only difference is that there is no human in the conversation — so any actions requiring approval will wait for your review.