소개
This skill empowers Claude to perform deep, systematic debugging by tracing errors backward through the execution chain. Instead of simply fixing where an error appears, it guides the AI to identify the original trigger point in the call stack, implement defense-in-depth validation at multiple layers, and use specialized techniques like instrumentation and test-pollution bisection. It is particularly effective for complex software architectures where errors manifest deep in execution, far from the source of invalid data or state.