What problem does it solve? Engineers often jump straight to editing code when facing ambiguous bugs, intermittent behavior, or performance regressions, which leads to wrong fixes and wasted effort. This Skill enforces an evidence-first diagnostic workflow so the root cause is identified before any change is made. ## Core Features & Use Cases - Evidence-Gathering Discipline: Separates observed symptoms from inferred causes and traces inputs, state transitions, ownership boundaries, and failure output. - Hypothesis Ranking: Ranks candidate causes by supporting evidence and cheap falsification value, stopping exploration once one credible mechanism explains the evidence. - Use Case: When a production service intermittently times out, use this Skill to trace the failure path, rank hypotheses such as connection pool exhaustion versus downstream latency, and report the proven cause before authorizing any fix. ## Quick Start Use investigate-first to diagnose this intermittent test failure and report the root cause with evidence before proposing any code changes.
