diagnose

To diagnose the root cause of flaky test failures, this skill constructs a reproducible test case from symptoms, logs, and environmental context to generate ranked hypotheses, a proven root cause, and remediation steps within a structured Findings Bundle.

The best way to perform evidence-driven root cause analysis for software regressions is to assess symptom signals and recent changes within a structured diagnosis loop, producing a Findings Bundle with evidence, ranked hypotheses, and a defined handoff to developer skills.

You create a structured handoff for reported software bugs by generating a Findings Bundle that specifies routing to the appropriate developer skill, including proven root cause analysis, remediation steps, and regression checks to accelerate reviews.

Yes, you can use this diagnosis approach for unclear software bugs because it avoids guesswork by guiding evidence-backed analysis, collecting artifacts, and forming testable hypotheses before outputting a defined remediation plan.

To generate a diagnosis report for a failing test, you need to provide a description of a reproducible failure, which the skill uses to capture symptom signals, environmental context, and recent changes to output a proven root-cause analysis.

You should use a structured diagnosis loop instead of standard debugging for hard bugs, performance regressions, and flaky failures, as it builds testable hypotheses and provides standardized output guidance with Findings, Plan, and Routing bundles for remediation.