design-system
Skill1 starsChanged 53 days agoMITView on GitHub
Create or update a site's DESIGN.md from reference brands, and implement it as tokens, fonts, and a favicon. References can be getdesign.md slugs, DESIGN.md files, or live URLs. Use it when the user wants to import a brand's style ("que se vea como Linear", "usa el DESIGN.md de Stripe"), change palette or typography, define motion or art direction, or when web-kit:new-site reaches phase 2.
193 other projects use a version of this.
Adopted by
Most recently taken up by
- lengrufeng2008/dsh-skill-librarytheir own version, first seen Nov 15, 2026
- dmoreq/precepttheir own version, first seen Nov 15, 2026
- kobiniv888-ai/Claude-codetheir own version, first seen Nov 15, 2026
- chuanglibs/Claude-Code-Game-Studios-Codextheir own version, first seen Nov 15, 2026
- jeremy-colard-dev/agent-communication-claudetheir own version, first seen Nov 15, 2026
- davision0102/fabriktheir own version, first seen Nov 15, 2026
Used with
Projects with it also use setup-engine, team-level and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
CoralGameSkillsetup-engine
Configure the project's game engine and version. Pins the engine in CLAUDE.md, detects knowledge gaps, and populates engine reference docs via WebSearch when the version is beyond the LLM's training data.
Also in 13 projects that use this
CoralGameSkillteam-level
Orchestrate level design team: level-designer + narrative-director + world-builder + art-director + systems-designer + qa-tester for complete area/level creation.
Also in 12 projects that use this
lilangMaxSkillcreate-stories
Break a single epic into implementable story files. Reads the epic, its GDD, governing ADRs, and control manifest. Each story embeds its GDD requirement TR-ID, ADR guidance, acceptance criteria, story type, and test evidence path. Run after /create-epics for each epic.
Also in 11 projects that use this
lilangMaxSkillstory-done
End-of-story completion review. Reads the story file, verifies each acceptance criterion against the implementation, checks for GDD/ADR deviations, prompts code review, updates story status to Complete, and surfaces the next ready story from the sprint.
Also in 11 projects that use this
- +2 more from CoralGame
- +6 more from lilangMax
Other versions193 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 193
- nextlevelbuilder/ui-ux-pro-max-skill★ 147ksame name
- getsentry/sentry★ 46ksame name
- anthropics/knowledge-work-plugins★ 36ksame name
- Donchitos/Claude-Code-Game-Studios★ 27ksame name
- ethereum/ethereum-org-website★ 6ksame name
- langwatch/langwatch★ 5ksame name
- joye61/pic-smaller★ 2.9ksame name
- daveebbelaar/bama★ 1.7ksame name
- reatom/reatom★ 1.4ksame name
- microsoft/power-platform-skills★ 1.2ksame name
183 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository7 other files it gives its agents.
- template/CLAUDE.mdCLAUDE.md
- assetsSkill
- design-loopSkill
- new-siteSkill
- publishSkill
- retroSkill
- site-copySkill
The file
We split this skill into the rules it gives an agent
11 rules. 0 are common, 5 are distinctive, and 6 are very specific to this project.
0
Common
Most of the 175 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
If the user has no references (usually they won't), don't ask for any.
- Distinctive: line 16
Ask with AskUserQuestion: the 3 directions plus "Mezclar" (they say what they like from each).
- Very specific: line 22
getdesign.md slug: run npx getdesign@latest add <slug>.
- Distinctive: line 26
Write down what each reference contributes.
- Distinctive: line 30
Use this structure. Every value must be concrete (hex, px, ms, font name).
- Distinctive: line 42
Don't copy logos, brand names, illustrations, or signature copy from the references.
- Very specific: line 46
Put the tokens in @theme in src/index.css: --color-*, --font-*, plus radius and shadow tokens if they're non-standard.
- Very specific: line 47
Put the fonts in index.html: add the Google Fonts preconnect + stylesheet link in place of the placeholder comment.
- Very specific: line 48
Write public/favicon.svg: a simple mark in the brand colors.
- Very specific: line 50
Run npm run build to verify.
- Very specific: line 52
When updating an existing system, change DESIGN.md first, then the tokens, then the components.
“Common” is measured: kept by at least 6 in 10 of the 175 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Design system
DESIGN.md is the contract every component, animation, and generated asset answers to. A site without one drifts toward AI defaults: indigo gradients, three generic cards, system fonts. The job is to synthesize an identity from references. Cloning one brand is not the job.
0. Let the user choose a direction
If the user has no references (usually they won't), don't ask for any. Propose them:
- From the brief, pick 3 contrasting directions that all suit the audience. For example: sober and technical, warm and close, bold and editorial. Base each one on 1–2 real references: getdesign.md slugs (
npx getdesign@latest list) or well-designed sites in the same niche. - Show each direction as a name, one sentence ("se siente como…"), what it would look like (colors, type, mood), and a screenshot of its main reference taken with the claude-in-chrome tools, or at least the link.
- Ask with AskUserQuestion: the 3 directions plus "Mezclar" (they say what they like from each). Put your recommendation first.
The chosen references go through step 1.
1. Collect references into docs/references/
- getdesign.md slug: run
npx getdesign@latest add <slug>. It writes./DESIGN.md. Immediately move it todocs/references/<slug>.md. - File the user gives you: copy it to
docs/references/<name>.md. - Live URL: open it with the claude-in-chrome tools. Take desktop and mobile screenshots. Read the computed styles of the body, the headings, a primary button, and a card:
getComputedStyleviajavascript_tool. Write down what you measured indocs/references/<domain>.md, using the same section structure as below.
Write down what each reference contributes. The brief in README.md §6 says, for example "structure and type from A, warmth and color from B".
2. Write DESIGN.md
Use this structure. Every value must be concrete (hex, px, ms, font name). A section that stays vague will be filled in by the model with defaults.
- Concept: one sentence of visual metaphor, plus the theme (light or dark). For example: "payment rails in a drafting studio".
- Colors: a table with Name, Value, Token (
--color-<name>), and Role. 6–10 colors. Name them by role or material (ink,paper,signal), never by hue. Mark exactly one accent color for conversion, and say where it's allowed. - Typography: display, body, and mono fonts from Google Fonts (verify the license is OFL, i.e. usable commercially), with weights. Add a type scale with size, line-height, and tracking per role (page title, section title, body, label).
- Spacing, radius, shadows: a scale for each.
- Components: exact recipes (classes/tokens) for buttons (primary, secondary), nav, cards, inputs, and section headers.
- Motion: easing curves, durations, what animates (entrances, hover, scroll) and what never does. Everything respects
prefers-reduced-motion. - Art direction (read by
web-kit:assets): subject or signature element (a distinctive one, not a cliché), lighting, palette as hex, abstraction level, composition, and forbidden clichés. - Do / Don't: 5 of each, specific to this site.
- Sources: which reference contributed what.
Don't copy logos, brand names, illustrations, or signature copy from the references. Palettes, type scales, and layout patterns are fair to learn from. Their identity isn't.
3. Implement
- Put the tokens in
@themeinsrc/index.css:--color-*,--font-*, plus radius and shadow tokens if they're non-standard. Update thesection-titleandpage-titleutilities to match the type scale, and thebodybase style. Remove the placeholder comment. - Put the fonts in
index.html: add the Google Fontspreconnect+ stylesheet link in place of the placeholder comment. Settheme-color. - Write
public/favicon.svg: a simple mark in the brand colors. Don't use a letter inside a circle. - Replace any template token names used in components (
muted,line,accent) with the new ones. Then check that nothing still references a removed token:grep -rn "text-muted\|border-line\|accent" src. - Run
npm run buildto verify.
When updating an existing system, change DESIGN.md first, then the tokens, then the components. A new color used by one component starts as a token.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

