reliability-reviewer

Detect missing reliability safeguards in microservice code by analyzing pull request diffs for unhandled I/O errors, unbounded retry loops, and missing timeouts on external calls. This review process identifies concrete cascading failure paths to prevent partial failures and production outages.

Cascading failures in backend services are caused by missing reliability safeguards like error swallowing, unbounded retry loops, and missing timeouts on external API and database calls. When dependencies fail, these code gaps cause partial failures to propagate and disrupt connected microservices.

Yes, you can review background jobs and async handlers for error swallowing issues. The analysis detects unhandled I/O errors and missing error handling on message queue boundaries specifically within code interacting with external dependencies like background processing systems.

Ensure your payment processing microservice degrades gracefully during gateway outages by identifying missing timeouts and retry limits on external payment gateway calls. This code review flags concrete cascading failure paths to guarantee systems degrade gracefully under failure conditions.

The best way to find reliability gaps in a pull request diff is to scan for missing error handling on I/O boundaries, unbounded retry loops, and missing timeouts. This approach assigns confidence scores to each finding based on how directly the gap is visible in the code.

No, this code review process does not flag speculative architectural issues. It assigns confidence scores to each finding based on how directly the reliability gap is visible in the code, avoiding speculative flags for unconfirmed architectural issues.