What problem does it solve? Fixing a bug is only half the job—without understanding why it slipped through, the same class of bug returns. This Skill guides a disciplined local bugfix workflow: reproduce the bug, identify the cause, apply a minimal fix, verify with type checks, lint, and tests, then perform root-cause analysis and implement structural prevention measures. ## Core Features & Use Cases - Guided Reproduction: Chooses the right reproduction method per bug type—failing tests, type checks, lint, dev server, or Storybook—and records the reproduction steps before fixing. - Minimal, Convention-Aware Fixes: Applies the smallest possible change while respecting project coding standards from CLAUDE.md and project configuration. - Root-Cause Analysis & Prevention: Uses git blame and git log to trace when the bug was introduced, then strengthens types, lint rules, compiler flags, tests, or architecture to prevent recurrence. - Use Case: A user reports a runtime error in a React component. The Skill reproduces it via Storybook, traces the cause through recent commits, fixes it, adds an exhaustive type check and a regression test, and reports the root cause with prevention measures. ## Quick Start Ask the AI to reproduce and fix the bug where submitting the form throws a type error, then verify with tests and explain the root cause.