investigate

Root cause analysis identifies why software failures occur through a structured, evidence-first investigation workflow. It prevents surface fixes by requiring evidence confirmation before implementing any code changes.

Debugging a stack trace requires a structured root-cause investigation process that examines repository evidence. This approach drives through investigate, analyze, hypothesize, and implement phases to resolve unexpected runtime behavior.

Incident response and error triage for runtime regressions follow a four-phase investigation workflow. It enforces an iron law of no fixes without first confirming the underlying root cause using repository inspection tools.

No, the root-cause-first rule explicitly prohibits implementing any fixes without confirming the root cause. This constraint reduces repeated outages and churn by ensuring every corrective path is justified by evidence.

Evidence-based fixes follow four sequential phases: investigate, analyze, hypothesize, and implement. You must use allowed tools for repository inspection to gather evidence and confirm the root cause before proposing the smallest safe fix.

You need root cause analysis when you see errors, 500s, stack traces, or behavior that regressed since it worked yesterday. It applies to debugging requests like resolving why software is broken and preventing repeated outages.