What problem does it solve? Color bugs in interfaces are usually system bugs: values picked in isolation, tokens borrowed because they looked right, and contrast pairs nobody measured. This Skill gives you a complete method for building and auditing a color system—ramps, semantic tokens, contrast measurement, and dark mode—so every color decision has an exact, verifiable answer. ## Core Features & Use Cases - Palette generation and structure: Build neutral, accent, and status ramps with even perceived lightness, constant hue, and role-mapped steps following Tailwind or Radix conventions. - Token naming and theming: Separate primitive tokens from semantic role tokens so dark mode, white-label themes, and increased-contrast variants work without touching components. - Contrast measurement and reporting: Measure foreground/background pairs against APCA or WCAG 2 thresholds and report failing pairs with severity, without silently repainting design decisions. - Use Case: You inherit a codebase with scattered hex values and a broken dark mode. Use this Skill to inventory every color literal, collapse near-duplicates, assign roles, rebuild the ramps, and produce a severity-ordered findings table. ## Quick Start Ask the AI to audit the color system in your project and report any contrast, token, or palette issues it finds.