cron-health

Scheduled task monitoring skill. Use when checking cron job health, diagnosing missed jobs, or setting up monitoring.

SKILL.md

GitHub

Monitor scheduled tasks and ensure nothing runs silently broken.

What to Track Per Job

Field Description
Name Job identifier
Schedule Cron expression
Last run Timestamp of most recent execution
Last status Success or failure
Duration How long it took
Output Key metrics or error message

Health Check Pattern

For each scheduled job:

  1. Has it run in the expected window? (compare schedule vs last_run)
  2. Did it succeed? (check status)
  3. Was the duration normal? (compare vs 7-day average)
  4. Did it produce expected output? (records_processed > 0)

Alert Conditions

  • Missed run — job didn't execute within 2× its scheduled interval
  • Consecutive failures — same job failed 3+ times in a row
  • Duration anomaly — took 3× longer than average
  • Zero output — ran successfully but processed 0 records (may indicate upstream data issue)

Daily Summary Format

Cron jobs (24h): X ok, Y failed
  Failed:
  • job_name: error description
  Slow:
  • job_name: took Xms (avg: Yms)

Safety: Before Restart

Before restarting any service, check:

  1. Are any cron jobs currently running?
  2. Are any scheduled to run in the next 5 minutes?
  3. If yes → defer the restart until the window is clear
  4. If no → safe to restart