The Cron Jobs page monitors the scheduled tasks defined in your OpenClaw agent configuration. Each job appears as a row you can expand to see its run history, duration trend, and the prompt it sends to the agent.
KPI strip
Five metrics at the top of the page give you a quick health check:
| Metric | Description |
|---|
| Jobs / enabled | Total job count and how many are currently enabled |
| Runs today | Number of job executions since midnight |
| Failures today | Runs that ended with error or failed status today — highlighted in red when non-zero |
| Failures (7d) | Error count over the past seven days |
| Next run | Countdown to the next scheduled execution, with the job name below it |
Job list
The job table has the following columns:
| Column | Description |
|---|
| Job | Job name and the agent ID it belongs to |
| Schedule | Cron expression (e.g. 0 9 * * 1) plus a human-readable translation (e.g. “weekly at Mon 09:00”) and timezone |
| Next run | Countdown until the next scheduled execution |
| Last run | Date and time of the most recent execution, its status icon, duration, and consecutive error count |
| Status | Badge showing the last run status |
| Enabled | Toggle to enable or disable the job without editing the config file |
×N in red next to the duration.
Enabling and disabling jobs
Click the toggle switch in the Enabled column to enable or disable a job. The change takes effect immediately. The toggle is disabled while the update is in flight, and an error message appears below the toggle if the update fails.
Expanded job detail
Click any job row to expand it. The expanded panel shows:
Prompt preview — the message payload that gets sent to the agent when the job fires. This is the text defined in the job’s payload.message field.
Run trend chart — a bar chart of the last 20 runs ordered chronologically (oldest left, newest right). Each bar represents one run’s duration in seconds. Green bars are successful runs; red bars are errors. A dashed reference line shows the average duration of successful runs.
Run history — up to 20 most recent runs, each displayed as a card. Each card shows:
- Status —
OK, ERROR, or RUNNING with an icon
- Run time — when the run started
- Duration — how long the run took
- Token usage — total tokens consumed, with input/output/cache-read/cache-write breakdown
- Model — the model that handled the run (abbreviated)
- Error message — shown in red when the run failed
- Summary — a brief description of what the run did (for successful runs)
- Session link — if the underlying session still exists in the Claw Lens database, a link to view it in the Sessions page
If a session has been pruned from local storage, the session link will not appear in the run card, even though the run record still exists.
Failure tracking
Claw Lens tracks consecutive errors per job. When a job fails repeatedly without a successful run in between, the count increments and is displayed in red next to the last run duration on the job row.
A non-zero consecutive error count means the job has not recovered on its own. Check the error message in the latest run card for details and fix the underlying issue before the agent accumulates more failed sessions.
Use cases
Monitoring automated agents — If you have agents that run on a schedule (e.g. daily summaries, periodic data pulls), the Cron Jobs page gives you a single place to confirm they ran, how long they took, and whether they succeeded.
Debugging failed scheduled tasks — When a job shows a red status, expand it to read the error message and link through to the Session Timeline to see exactly which tool call failed and what the raw output was.
Capacity planning — The run duration trend and token usage per run help you understand how resource-intensive your scheduled agents are and whether they are growing over time.
Empty state
If no jobs appear, Claw Lens has not found any scheduled tasks in your OpenClaw configuration. This means you have not defined any cron jobs yet. 