What problem does it solve? Codebases often accumulate inconsistent colors, arbitrary spacing, and generic AI-generated styling that makes interfaces look unpolished, and teams lack a systematic way to define or audit their visual language. ## Core Features & Use Cases - Design System Generation: Scans CSS, Tailwind, and styled-components to extract existing patterns, then produces design-tokens.json, DESIGN.md with rationale, and a self-contained HTML preview page. - Visual Audit: Scores a UI across 10 dimensions including color consistency, typography hierarchy, spacing rhythm, accessibility, and polish, with fixes citing exact file:line locations. - AI Slop Detection: Flags generic AI-generated design patterns such as gratuitous gradients, purple-to-blue defaults, and purposeless glass morphism. - Use Case: Before a redesign of a SaaS app, run an audit against localhost pages to get scored feedback on every visual dimension, then generate a cohesive token set to guide the new implementation. ## Quick Start Ask the AI to generate a minimal design system with earth-tone palette for your project, or to audit your running app at a local URL for visual consistency.