Try it
Add the skill to a bot, then ask your Chief of Staff:
“Use the Background Jobs Patterns skill on this: [describe the job, or paste your notes].”
For when a queue broker (Redis, RabbitMQ) goes down and recovers, all enqueued jobs become visible simultaneously, and all retrying jobs retry at the same moment. This can overload downstream dependencies (databases, external APIs).
What it covers
- Background Job Architecture Assessment