What problem does it solve? Vague code review feedback like "this feels messy" gives authors nothing to act on or look up. This Skill turns subjective impressions into named, citable findings by mapping observations to Martin Fowler's code smells and refactoring catalog, Google's code review standard, and official React, Next.js, and TypeScript guidance. ## Core Features & Use Cases - Named diagnoses and fixes: Maps observations to exact smell names (Long Function, Feature Envy) paired with catalog refactorings (Extract Function, Introduce Parameter Object) so every finding is traceable to a first-party source. - Bounded severity discipline: Applies Google's Major/Minor/Nit scale with the Nit: prefix convention, so reviews distinguish blocking issues from polish and avoid noise. - Framework-specific checks: Covers the thirteen React useEffect anti-patterns, Next.js Server/Client Component boundary rules, and TypeScript do's and don'ts for App Router codebases. - Use Case: When reviewing a Next.js pull request, load this Skill to produce structured findings with file locations, current and suggested code, and citations like "Fowler, Long Parameter List -> Introduce Parameter Object" that authors can verify independently. ## Quick Start Review the code in 15-prototype and produce a findings report using named code smells, refactorings, and severity levels.
