What problem does it solve? Debugging often devolves into random code changes until the bug seems to disappear, masking root causes and wasting hours. This Skill enforces a disciplined, hypothesis-driven methodology that finds the actual root cause instead of treating symptoms. ## Core Features & Use Cases - Five-Step Method: Reproduce the bug, form a falsifiable hypothesis, test one variable at a time, isolate the root cause with Five Whys, then fix and verify with a regression test. - Isolation Techniques: Choose from binary search (git bisect), differential debugging, trace logging, rubber duck explanation, and minimal reproduction based on what you know about the bug. - Cognitive Bias Guardrails: Built-in countermeasures for confirmation bias, anchoring, recency bias, and availability bias that derail investigations. - Time-Boxing Rules: Escalation triggers at 15 minutes, 30 minutes, 1 hour, and 2 hours to force approach changes before you get stuck. - Use Case: A production bug where order data appears stale. Instead of guessing, you reproduce it, hypothesize that cache invalidation is missing after order creation, confirm with a single log line, trace the root cause to missing integration test coverage, and ship both the fix and the test. ## Quick Start Ask the AI to apply the systematic debugging method to investigate why your endpoint returns stale data, starting with reproduction steps and a falsifiable hypothesis.
