What it does
generate-design-md creates and maintains DESIGN.md files — machine-readable design system documents that AI agents can understand and follow. Each file combines YAML design tokens (colors, typography, spacing) with markdown rationale.
How it works
The skill offers two generation paths: describe your design vision and let Claude generate tokens, or scan your existing codebase to extract design signals automatically. After generation, it validates your file against Google’s linting rules, checking for broken references, WCAG contrast compliance, and proper structure. You can export tokens to Tailwind CSS or W3C DTCG format.
Use cases
- Design system documentation: Create single-source-of-truth design files that both humans and AI agents read
- Design handoff: Give design agents consistent visual rules so generated screens match your brand
- Token management: Maintain colors, typography, and spacing in one validated format
- Scaling design: Ensure consistency across AI-generated interfaces and components
Who benefits
Product designers building design systems, design leaders standardizing brand assets, and teams using AI agents for UI generation.