What problem does it solve? Debugging under pressure often turns into guess-and-check: quick fixes that don't hold, stacked patches that add new variables, and hours lost investigating the wrong component. This Skill enforces a root-cause-first process so fixes address the actual source of a bug instead of its symptom. ## Core Features & Use Cases - Four-phase debugging process: reproduce the failure, trace the bad value backward to its origin, state one testable hypothesis, then write a failing test before applying a single targeted fix. - Architecture escalation rule: after three failed fix attempts, stop patching and question the design, with explicit red flags and rationalization tables to catch thrashing early. - Supporting techniques and tooling: reference guides for root-cause tracing, defense-in-depth validation, and condition-based waiting for flaky tests, plus a script that bisects test files to find which one pollutes shared state. - Use Case: A CI suite fails intermittently and a stray .git directory keeps appearing. Run the find-polluter script to identify the offending test file, trace where the empty path value originated, fix it at the source, and add validation so it cannot recur. ## Quick Start Use the systematic-debugging skill to investigate why this test fails intermittently before proposing any fix.
