Cron Watchdog Self-Heal

Automatically detects and resumes cron jobs that have stopped, become disabled,
paused, or entered a "completed" state when they should still be running.

When to Use

• User reports "cron jobs stopped firing"
• User asks about cron reliability or uptime
• Routine: runs every 5 minutes via dedicated self-heal cron job

Procedure

1. List all jobs via cronjob(action='list')
2. Inspect each job's state. A job needs healing if:
   • enabled: false AND state is not intentionally paused (paused_reason is not set)
   • state: "completed" when it should be recurring
   • last_status indicates an error
3. Re-enable by calling cronjob(action='update', job_id='...') — re-posting the same config with enabled: true
4. Log the fix — note which jobs were healed and why
5. Deliver a quiet report in the chat (one line per healed job)

Jobs to watch

Only recurring cron jobs (repeat: forever) should be auto-healed. One-shot or intentionally paused jobs are skipped.

Pitfalls

• Do NOT heal intentionally paused jobs (check paused_reason).
• Do NOT heal one-shot jobs.
• If a job keeps failing immediately after heal, report it — don't loop-heal.
• Rate-limit: only report actual changes, not "all clear" every tick.

Verification

cronjob(action='list')  # All jobs should show enabled: true, state: scheduled