One of OpenClaw’s most powerful features is its built-in cron job scheduler. Instead of manually asking your agent to perform repetitive tasks, you can schedule them to run automatically — morning briefings, inbox checks, data monitoring, and more.
How OpenClaw Cron Jobs Work
Cron jobs run inside the Gateway process, not the AI model itself. Jobs are persisted at ~/.openclaw/cron/jobs.json, so they survive restarts. You can choose between two execution modes:
- Main session: The job runs within your existing conversation context.
- Isolated session: Each run starts a fresh session with no prior conversation carry-over — ideal for independent tasks.
Creating Your First Cron Job
One-Shot Reminder
openclaw cron add \
--name "Reminder" \
--at "2026-03-08T09:00:00Z" \
--session main \
--system-event "Reminder: review the weekly report" \
--wake now \
--delete-after-run
Recurring Morning Briefing
openclaw cron add \
--name "Morning brief" \
--cron "0 7 * * *" \
--tz "Asia/Hong_Kong" \
--session isolated \
--message "Summarize overnight emails, calendar for today, and top 3 AI news stories." \
--announce \
--channel telegram
Useful Cron Job Ideas
- Morning briefing (7 AM): Summarize emails, calendar, and news.
- End-of-day review (6 PM): Recap tasks completed and flag pending items.
- Competitor monitoring (every 4 hours): Check competitor websites for pricing or feature changes.
- Memory flush (11 PM): Run a pre-sleep memory consolidation to prevent context loss overnight.
- Weekly report (Monday 9 AM): Generate a summary of the past week’s activities.
Managing Cron Jobs
# List all jobs
openclaw cron list
# Run a job manually
openclaw cron run <job-id>
# View past runs
openclaw cron runs --id <job-id>
# Delete a job
openclaw cron delete <job-id>
Pro tip: Schedule resource-intensive cron jobs during off-peak hours (late night) to avoid hitting API rate limits during your active working hours. Use Haiku or smaller models for cron tasks to keep costs down.
Explore more tips in our OpenClaw User Tips series.


