systematic-debugging

To find the root cause of flaky test failures, you must complete a four-phase workflow: investigation, pattern analysis, single-hypothesis testing, and root-cause implementation. This enforces a strict mandate against symptom patching before verifying the actual issue.

Debugging production bugs under time pressure requires a pressure-resistant approach that blocks shortcuts. You should trace data flow backward, test a single hypothesis, and mandate re-analysis or architectural review if repeated fixes fail to resolve the unexpected behavior.

To stop repeated fixes from failing on build failures or integration problems, enforce re-analysis after up to two failed attempts. If fixes still fail, you must escalate to an architectural review to question the foundational design.

You should perform an architecture review during root cause investigation after repeated fixes fail. The systematic debugging process mandates architectural questioning when implementation attempts exceed two failures, ensuring deeper structural issues are addressed.

Hypothesis testing for performance issues works by isolating a single proposed root cause and validating it with a minimal failing test. You trace data flow and apply defense-in-depth validation before implementing any actual code fixes.

Yes, backward tracing and condition-based waiting are companion techniques used to reliably reproduce test failures. They support the investigation phase by tracing data flow backwards and waiting for specific states before proposing a root-cause fix.