What problem does it solve? Retrofitting a mature design system onto tokens is risky without knowing its true size: how many hard-coded colors live in the codebase, how many Figma variables and bindings exist, and how much is already semantic. This Skill measures both sides before anything is built or changed, so the retrofit effort is right-sized and nothing is blindly renamed or destroyed. ## Core Features & Use Cases - Code surface measurement: Runs a configurable grep scaffold that counts SCSS color variables, Tailwind color classes, JS color usages, raw hex/rgba literals, and SVG fills, with tunable patterns per repo. - Verified Figma inventory: Independently reads variables, consuming bindings, text styles, effect styles, and modes, re-reading suspicious empty results instead of guessing. - Semantic percentage scoring: Computes a 0-100 percentSemantic score that decides whether the retrofit is a rename-in-place cleanup or a rewrite, and records documentation debt for planning. - Use Case: A team with a five-year-old codebase and a drifted Figma file runs the audit, learns they are ~90% semantic with 1,200 bindings, and confidently plans a rename-and-cleanup retrofit instead of a rebuild. ## Quick Start Audit my existing design system by measuring the color usage in my repo and inventorying the variables, styles, and bindings in my connected Figma file.
