What problem does it solve? Debugging often devolves into guesswork: shotgun changes, symptom fixes, and unverified hypotheses. This Skill enforces a disciplined investigation process that traces the full causal chain from trigger to symptom before any fix is proposed, so bugs are actually resolved instead of masked. ## Core Features & Use Cases - Systematic Root-Cause Investigation: Reproduces the bug, verifies environment sanity, traces the code path backward to where valid state first became invalid, and gates the fix on a complete causal chain with testable predictions. - Issue Tracker Integration: Fetches full issue threads from GitHub via gh, Linear, or Jira, including all comments, to extract reproduction steps and prior failed attempts. - Test-First Fixing with Safe Handoff: Writes a failing test before the minimal fix, runs regression suites, then commits and opens a PR with tracker auto-close syntax (e.g., Fixes #N). - Use Case: A developer pastes a stack trace or a GitHub issue URL. The Skill reproduces the failure, forms ranked hypotheses with falsifiable predictions, confirms the root cause, and delivers a tested fix on a new branch with a PR ready for review. ## Quick Start Ask the AI to debug the failing test or pasted error message and find the root cause before fixing it.
