Cron Whisperer
Document and monitor the cron jobs nobody understands
Audits all cron jobs on a server, documents what each one does, wraps them with monitoring and alerting, and migrates critical ones to a proper scheduler if needed. No more "30 cron jobs and nobody knows what half of them do."
INGREDIENTS
PROMPT
Create a skill called "Cron Whisperer". Audit all cron jobs on this system: (1) List every crontab entry for every user. (2) For each entry, read the script it runs and generate a plain-English description of what it does, how long it typically takes, and what resources it uses. (3) Identify dependencies between jobs (shared files, database tables, output/input relationships). (4) Check recent execution history for failures (exit codes, error logs, missing output files). (5) Generate a documentation page for each job. (6) Wrap each job with a monitoring script that: checks for failure, retries once, and sends an alert to Slack or Telegram on failure. (7) If any jobs are critical enough to warrant a real scheduler, generate an Airflow DAG or GitHub Actions workflow as a migration path.
How It Works
Every data team has a server with mysterious cron jobs that a former team
member set up. They run critical ETL processes with no monitoring, no retry
logic, and no documentation. When one silently fails, you discover it days
later. This skill audits the chaos and adds safety nets.
What You Get
- Full audit of all cron jobs with plain-English descriptions of what each does
- Dependency mapping (which jobs depend on others completing first)
- Health monitoring with Slack/Telegram alerts on failure
- Retry logic wrappers for flaky scripts
- Documentation for each job (purpose, owner, dependencies, failure impact)
- Migration plan for moving critical jobs to a proper scheduler
Setup Steps
- Ask your Claw to create a "Cron Whisperer" skill with the prompt below
- Point it at the server(s) with cron jobs
- Review the audit report and fill in any missing context
- Enable monitoring wrappers on critical jobs
Tips
- The audit alone is valuable — many teams have never inventoried their cron jobs
- Start monitoring before migrating — know what breaks before you change things
- The dependency mapping often reveals jobs that should run in sequence but don't
- For small teams, wrapping cron with monitoring is often better than migrating to Airflow