performance-review

To detect performance bottlenecks during code review, systematically audit source code for algorithmic inefficiencies, N+1 queries, and blocking I/O operations that cause system degradation at scale.

Common performance anti-patterns causing latency include unbounded memory usage leading to OOM errors, event-loop stalls, database inefficiencies lacking proper indexes, and algorithmic complexity issues.

Evaluate concurrency bounds and blocking I/O operations by systematically assessing growth variables, query shapes, and event-loop stalls to ensure production stability in resource-constrained environments.

Yes, this performance review process is explicitly designed for async code in resource-constrained environments and high-traffic services, evaluating event-loop stalls and concurrency bounds.

Distinguish critical performance blockers from minor suggestions using a structured severity rubric that calibrates findings based on the blast radius and call frequency of the identified bottleneck.

The best way to identify unbounded memory allocation and scalability issues is running a systematic performance hunt protocol on the current diff to detect loops lacking memory bounds and algorithmic complexity.