ralph-wiggum_skill

Overview

This skill is a surgical debugger and code optimizer for JavaScript projects that focuses exclusively on fixing existing logic, not adding features. It enforces a forensic-first workflow: reproduce the failure, identify the root cause, and apply minimal, high-fidelity fixes that defend architecture and test integrity.

How this skill works

The skill runs an orchestrated harness to reproduce failing tests and trace bad values back to their origin. Fix attempts go through a persistent loop: write a failing test, implement the smallest corrective change, run reflection checks, and iterate up to safe circuit-breaker limits. Every patch must embrace modularity, clear naming, and pass the reflection checklist before being accepted.

When to use it

• When automated tests fail and you need a focused root-cause investigation.
• When recurring bugs reappear despite symptom-level patches.
• When code smells suggest architectural drift or fragile coupling.
• Before merging hotfixes that could increase technical debt.
• When you need regression-proof fixes with test coverage.

Best practices

• Always start with Phase 1: reproduce the failure and trace the bad value back to its origin.
• Write a failing unit or integration test that precisely captures the defect before changing code.
• Favor surgical, minimal changes that fix the architectural flaw rather than hiding symptoms.
• Run the reflection loop to validate edge cases, input validation, and security concerns.
• Break fixes into single-responsibility functions and add tests for each slice.
• Use the harness loop conservatively; stop and re-evaluate after repeated identical failures.

Example use cases

• Eliminate a flaky function by tracing intermittent bad inputs and hardening validation.
• Refactor a brittle helper into smaller, well-named functions with targeted tests.
• Fix a regression introduced by a recent commit and produce a rollback plan if needed.
• Replace an incorrect algorithm with a clearer, tested alternative while preserving API behavior.
• Reduce blast radius by isolating stateful logic and adding defensive tests.