jobs

Periodic jobs are common (cleanup, digests, rollups), and it’s easy to accidentally run them on every instance in production. In local dev, in-process node-cron is fine. In production, I either run a dedicated worker or add a guard so only one instanc