code-smell

What problem does it solve? After standard code review confirms code is correct, there is no systematic check for whether the code is over-engineered, contains dead code, or carries deprecated leftovers. This Skill fills that gap with a taste-focused audit that catches unnecessary abstractions, unreachable branches, and residual clutter before tests run. ## Core Features & Use Cases - Eight-Category Checklist: Audits useless code, over-decoration, over-design, meaningless code, unreachable code, deprecated residue, architecture over-engineering, and hardcoding versus over-engineering balance. - Chained Review Workflow: Runs after a standard review-fix pass and before tests, producing a structured verifier report with verdict, evidence, over-engineering warnings, and priority ordering. - Use Case: After refactoring a Spring service, run this audit to catch a single-implementation interface, an unused import, a redundant try/catch that only rethrows, and a commented-out dead code block, then receive a file-and-line evidence list with suggested fixes. ## Quick Start Ask the AI to audit the recently changed files against the code-smell eight-item checklist and report any over-engineering or dead code with file and line evidence.