What problem does it solve? Bug reports often arrive with vague symptoms and no confirmed root cause, leading to premature fixes, redundant tests, and regressions that slip through existing coverage. This Skill turns a defect report into a structured root-cause analysis with reproduction evidence and a ready-to-establish verification plan, without editing any code. ## Core Features & Use Cases - Reproduction-First Investigation: Builds the fastest feedback loop to reproduce the defect before any fix is proposed, and stops with a clear report when reproduction is impossible. - Root Cause Confirmation: Traces the exact code path and condition producing the wrong behavior, separates confirmed observations from assumptions, and identifies invariants the fix must preserve. - Focused Verification Strategy: Recommends reusing, updating, or adding one focused regression at the lowest appropriate test level, with distinct handling for production defects versus test-infrastructure defects. - Use Case: A GitHub issue reports that a CLI flag is ignored. The Skill gathers the issue evidence, reproduces the failure with the narrowest command, confirms the root cause in the dispatch layer, and outputs a BugAnalysis recommending one integration-level regression. ## Quick Start Ask the AI to investigate the bug described in a GitHub issue or user report and produce a BugAnalysis with confirmed root cause and a proposed regression strategy.
