This repository is a curated collection of design system specifications for over 80 prominent brands, formatted as DESIGN.md files. Each file contains a structured YAML frontmatter defining design tokens (colors, typography, spacing, components) and extensive prose documentation covering visual theme, layout principles, elevation, responsive behavior, and iteration guidelines. The primary purpose is to provide AI coding agents (e.g., Claude, Cursor) with machine-readable design language definitions so they can generate UI that is visually consistent with a specific brand's identity. The repository also serves as a top-of-funnel marketing asset for VoltAgent's commercial getdesign.md service, which offers paid priority generation of new design analyses. The data lineage involves automated extraction from public marketing websites using tools like the dembrandt frequency analyzer and manual pixel sampling, followed by synthesis into a standardized template. The resulting specifications include exact hex values, font stacks, spacing scales, and component definitions with token references. While the repository is not a software application, it represents a structured dataset of design knowledge that can be consumed by both humans and automated tooling. Notable patterns include a consistent 9-section document structure, a token reference system using {token.ref} syntax, and a set of 10 standardized example components (ex-*) that are re-skinned across brands for comparison. The repository is actively maintained with a large community (122k stars, 13k forks), though contribution is centralized and the format is still in alpha. Known gaps include missing hover states, mobile screenshots, and form validation patterns. The design analyses are third-party interpretations and may drift from official brand assets over time.