What problem does it solve? Shipped interfaces often read as generically AI-generated, with predictable palettes, missing interaction states, contrast failures, and no hierarchy. This Skill provides a deliberate audit pass over an existing UI to score what makes it look generated and drive scoped refinements instead of vague redesigns. ## Core Features & Use Cases - AI-slop detection: Runs a two-altitude category-reflex test to catch first-order and second-order generic design choices, plus a register classification (brand vs product) that sets the right quality bar. - Heuristic critique and technical audit: Applies first-impression and squint tests, interaction-state coverage checks, contrast ratios (4.5:1 body text), accessibility, responsive, and performance audits with per-dimension N/10 scoring. - Scoped refine modes: Drives focused edits via modes like polish, harden, distill, bolder, quieter, clarify, layout, and typeset rather than wholesale rewrites. - Use Case: After shipping a dashboard, ask for a UI review to get scored findings on hierarchy, missing empty/error states, and contrast failures, then run a clarify and layout refine to fix them. ## Quick Start Review the UI I just built for AI-slop patterns, interaction-state gaps, and contrast issues, then propose a scoped refine.
