Core answer
Start here. This is the direct spoken answer to practice first.
Overview
A production-readiness review is a risk assessment tied to this service's critical behavior, not a generic checklist ceremony.
I start with the service's users, critical operations, dependencies, data ownership, and expected load. Then I review request and transaction boundaries, timeouts, idempotency, authentication and authorization, configuration, secret handling, health probes, graceful shutdown, and failure behavior. I require evidence from tests, load results, deployment configuration, and operational dashboards rather than accepting that an annotation or dependency is present.