correctness-reviewer

To detect logic bugs during a code review, trace execution paths with concrete input values to identify off-by-one errors, incorrect loop bounds, and boundary condition mistakes without needing runtime context.

Swallowed errors and invalid state transitions occur when error propagation fails, causing misrouted error handlers or fallback values that mask failures and leave systems in inconsistent, half-updated conditions.

Yes, you can validate null safety and error propagation from code alone by tracing execution paths to flag unhandled null values and misrouted error handlers without relying on runtime context or test suites.

Standard code reviews often miss boundary condition mistakes and pagination miscalculations, requiring automated correctness validation that traces execution paths with concrete inputs to catch these subtle logic errors.

The best way to audit error propagation failures is to trace execution paths to catch swallowed errors and fallback values that mask failures, preventing unhandled null values from crashing upstream callers.

You need automated correctness validation for state management bugs when standard reviews miss partial update issues that leave distributed systems in inconsistent conditions, requiring execution path tracing to identify reproducible bugs.