A collection of DESIGN.md files analysis by popular brand design systems. Drop one into your project and let coding agents generate a matching UI.
100.5k
Stars
11.6k
Forks
302
Open issues
4
Contributors
AI Analysis
A curated collection of DESIGN.md files — plain-text design system documents derived from analysis of popular brand design systems and developer-focused websites — intended to be dropped into a project so that AI coding agents can generate visually consistent UI matching a chosen design language. It serves the specific use case of AI-assisted UI generation (vibe coding), particularly for developers and designers who want their agent-built interfaces to match a recognizable design system witho...
Inferred from signals mentioned in the README (tests, CI, type safety) — not a review of the actual code.
AI's overall editorial judgment — not an average of the bars above, can weigh other factors too.
Curated DESIGN.md files let AI coding agents replicate brand UI styles instantly, no Figma required
awesome-design-md is a growing collection of DESIGN.md files — plain-text design system documents derived from analyzing real developer-focused websites (Vercel, Cursor, Anthropic, etc.). The concept, introduced by Google Stitch, lets AI coding agents read a single markdown file and generate UI that matches a brand's visual language. The target audience is developers, vibe-coders, and AI-assisted builders who want consistent UI without manually documenting design tokens or exporting from design tools. With 73 files and active additions, it functions as a design-system shortcut layer for the AI-coding workflow.
Created March 31 2026, timed closely with Google Stitch's introduction of the DESIGN.md concept. Appears to be a rapid ecosystem response to an emerging standard, paralleling how AGENTS.md repositories emerged after coding agent conventions were popularized.
91,717 stars in under 3 months is unusually fast growth, driven by the viral appeal of the AGENTS.md/DESIGN.md convention wave, strong alignment with the AI-assisted coding trend, and a simple zero-friction value proposition: copy one file, get better AI-generated UI. Still gaining ~1,744 stars per week as of the evaluation date, suggesting the momentum is sustained rather than a single viral spike.
The repo claims 'Ranked #150 globally on GitHub' in the README itself. The companion domain getdesign.md with a request/request endpoint and a commercial sponsorship model suggests real usage beyond passive starring. Direct adoption evidence from end-user projects is not independently verifiable from available metadata, but the forks (10,888) indicate active copying into real projects rather than purely passive interest.
Appears to be a static content repository — likely a flat or categorized collection of markdown files, with no runtime code. The README links to a companion site (getdesign.md) suggesting some tooling exists outside the repo itself. Likely no build pipeline beyond README generation and badge automation.
not documented in README — content repositories of this type typically have no automated tests; quality control appears to be editorial/curatorial rather than automated.
Last push June 16 2026, four days before the evaluation date. Active commit cadence implied by the 'Last update' badge and ongoing additions to the collection. Maintenance appears healthy and ongoing, not just passive hosting.
ADOPT IF: you are building UI with AI coding agents (Cursor, Claude, Copilot, Stitch) and want to match a specific developer brand's visual language without manually reverse-engineering their design tokens. AVOID IF: you need a production design system with versioned tokens, component libraries, or accessibility guarantees — this is a prompt-assistance artifact, not an engineering-grade design system. MONITOR IF: you are evaluating whether the DESIGN.md convention gains enough tool-native support (e.g., in Cursor, Windsurf, Claude) to become a de facto standard worth formally adopting in your org's AI workflow policies.
Independent dimensions
Mainstream potential
7/10
Technical importance
5/10
Adoption evidence
4/10
- The DESIGN.md format is a Google Stitch proposal, not a ratified standard — if Stitch is discontinued or the format pivots, the entire convention could lose tooling support.
- Files are hand-analyzed snapshots of brand design systems that change over time; the collection may become stale as brands evolve their visual identity.
- Quality of individual DESIGN.md files is hard to verify independently — some may produce poor AI outputs due to incomplete or inaccurate design token extraction.
- The commercial layer (getdesign.md, sponsorships, paid private requests) introduces incentive to prioritize popular brands over comprehensiveness or accuracy.
- Rapid star growth in very young repos sometimes reflects social sharing rather than deep usage; sustained fork-and-use behavior is a better signal and will take more time to validate.
Likely to become a de facto reference collection if DESIGN.md gains broad AI-tool support. Growth will plateau once the convention either standardizes or fades; the companion commercial site suggests the maintainers are positioning for the former.
Newsletter
Get analyses like this every Monday
Free weekly digest of the most interesting open-source discoveries.
Languages
No language breakdown available.
Information
- Website
- https://getdesign.md/
- License
- MIT
- Last updated
- 3w ago
- Created
- 3mo ago
- Analyzed with
- anthropic/claude-sonnet-4-6
Stars over time
Contributors over time
Top 100 contributors only — repos with more will plateau at 100.
Open issues
Top contributors
Recent releases
No releases published yet.
Similar repos
google-labs-code/design.md
DESIGN.md is a format specification that enables AI coding agents to understand...
superdesigndev/superdesign
SuperDesign is an AI design agent that generates UI mockups, components, and...
VoltAgent/awesome-agent-skills
Awesome Agent Skills is a curated collection of 1,500+ skills for AI agent...
| Repository | Stars | Week Δ | Language | Score | Updated |
|---|---|---|---|---|---|
|
VoltAgent/awesome-design-md
this
|
100.5k | +4.6k | — | 7/10 | 3w ago |
|
google-labs-code/design.md
|
25.5k | — | TypeScript | 8/10 | 1w ago |
 agentsmd/agents.md
|
22.9k | — | TypeScript | 7/10 | 4mo ago |
|
superdesigndev/superdesign
|
6.6k | — | TypeScript | 6/10 | 2w ago |
|
VoltAgent/awesome-agent-skills
|
27.7k | — | — | 8/10 | 1w ago |
 nexu-io/open-design
|
77k | — | TypeScript | 8/10 | -48 min ago |
Curates design inspiration resources broadly; not targeted at AI agent consumption or markdown-first workflows. Different audience and purpose.
Lists design tools and software; also not machine-readable by AI agents. Addresses designers choosing tooling rather than developers prompting agents.
Direct conceptual sibling — AGENTS.md for coding behavior, DESIGN.md for visual behavior. They address complementary concerns within the same AI-agent conventions ecosystem.
A TypeScript-based design tool, likely more interactive/programmatic. Serves a different modality (tool vs. static reference document) for a partially overlapping audience.
Covers software architecture patterns, not visual design systems. Serves a fundamentally different problem despite the 'design' overlap in the name.
