error-resilience-review

Audit code changes by inventorying external interactions in a diff, then classifying risks across error handling, retries, timeouts, idempotency, and graceful degradation to return a prioritized set of required fixes.

Graceful degradation turns outbound call failures into safe, degraded experiences by enforcing bounded timeouts, controlled retry budgets with backoff, and clear rollback paths to prevent cascading outage storms across services.

Add idempotency keys and retry budgets to outbound HTTP calls by enforcing bounded timeouts alongside controlled retry and backoff strategies, ensuring robust failure handling and preventing duplicate processing.

Resilience review applies to changes affecting scheduled tasks, messaging queues, databases, and any feature with external dependencies, evaluating timeout boundaries, observability hooks, and degradation strategies.

Use bounded timeouts and backoff for external API dependencies whenever you ship changes affecting outbound calls, ensuring failures are handled robustly with clear observability hooks and degradation paths.