What problem does it solve?
This Skill helps you debug broken software systematically instead of guessing. It is designed for situations where tests start failing, builds break, output is wrong, regressions return, or a recent change introduced an unfamiliar error.
Core Features & Use Cases
- Disciplined root-cause analysis: Enforces a reproduce → trace upstream → failing test → minimal fix workflow so fixes address the source of the bug rather than hiding the symptom.
- Safer debugging decisions: Adds guardrails such as rollback-first behavior, one-hypothesis-at-a-time investigation, hard stops after repeated failed attempts, and escalation guidance for flakes or multi-component failures.
- Structured debugging artifacts: Provides reusable templates for a debug report and a hypothesis ledger, plus references for root-cause tracing and intermittent test triage.
- Use cases: Debug a regression introduced by a recent commit, isolate a flaky test caused by shared state, trace a UI failure through browser evidence, or identify which layer is failing in a cross-service system.
Quick Start
Ask the AI to use the debug-issue skill to reproduce the failure with one deterministic command, trace the bug upstream to its root cause, write a failing regression test, and apply the smallest safe fix.