Systematic Diagnosis

Debug intermittent failures by building a deterministic reproduction harness that reliably triggers the symptom before proposing root cause analysis or fixes. This establishes a fast feedback loop to verify the exact environment-specific conditions causing the failure.

Find the root cause of a performance regression by applying falsifiable hypothesis testing with one-variable instrumentation. You add minimal tagged probes to validate assumptions, generating ranked explanations with clear predictions to isolate the exact bottleneck.

Start diagnosing a complex bug by providing your failure description along with the stack trace and logs to create a deterministic reproduction case. This reproduction-first constraint ensures you verify the symptom consistently before generating any testable explanations or fixes.

Hypothesis testing works for hard bugs by producing ranked, falsifiable explanations with clear predictions instead of guesses. You validate these assumptions through targeted instrumentation, adding minimal tagged probes to confirm or rule out specific failure mechanisms.

Write a regression test at the correct seam to verify your bug fix, then confirm the test passes and the instability is removed. This verified regression proves the fix resolves the exact issue while preventing future reintroduction of the failure.

Avoid systematic diagnosis when you cannot establish a deterministic reproduction of the failure, because the methodology strictly requires a reproducible test case before proposing explanations. Intermittent issues lacking measurable verification are unsuitable until a consistent harness is built.