reliability-reviewer

To detect missing error handling and timeouts in code reviews, analyze HTTP I/O, database queries, and async handlers for missing try/catch blocks, unbounded retry loops, and missing explicit timeout configurations.

A retry storm occurs when unbounded retries overload a system. Prevent retry storms by flagging retry loops during code review and ensuring they implement max attempt limits, exponential backoff, and jitter.

Identify cascading failure risks in async handlers by reviewing external calls for missing explicit timeouts that can hang and exhaust resources, and by checking for catch-and-ignore patterns that propagate misleading silent failures.

Yes, reliability review works for background jobs and database queries by applying failure mode detection across HTTP I/O, file operations, and message queues to ensure proper error handling and explicit timeout configurations are present.

Common error swallowing patterns in asynchronous operations include catch-and-ignore blocks that silently fail and propagate misleading defaults, masking true failure propagation paths and preventing graceful degradation.