ui-design
Skill213 starsChanged 3 days agoMITView on GitHub
Unify product UI design, frontend implementation, and verification. Use when asked to design, redesign, build, or fix a web, desktop, or mobile interface; a page, dashboard, form, component, design system, responsive layout, accessibility, interface animation, or editable hero/banner/infographic within a product surface. Combine visual decisions with implementation. Do not use for backend-only work, generic architecture, standalone artwork or slide decks, or browser-test-only requests.
We split this skill into the rules it gives an agent
19 rules. 0 are common, 12 are distinctive, and 7 are very specific to this project.
0
Common
Most of the 49 edited versions we read keep these.
12
Distinctive
This author's own choices. Worth a look.
7
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Use one product-surface loop.
- Very specific: line 28
Read ../ui-ux-pro-max/SKILL.md for a new surface, substantial redesign, or a visual/UX decision that needs evidence.
- Very specific: line 33
Read ../frontend-design/SKILL.md while changing an HTML, browser, or UXP frontend.
- Very specific: line 37
Read ../motion-framer/SKILL.md only when the changed surface is React/JS and the project already uses motion/framer-motion, or the user has …
- Distinctive: line 53
Do not load all four by default.
- Distinctive: line 70
Verify the changed flow at the narrowest real boundary: project checks first, then live behavior appropriate to the target when runnable …
- Distinctive: line 81
Keep an established design system unless the task is explicitly a redesign.
- Distinctive: line 82
Keep UI Pro Max as retrieval evidence, not as an authority over the product's brand, accessibility contract, or existing UX research.
- Distinctive: line 84
Treat motion as progressive enhancement.
- Distinctive: line 86
Do not broaden a UI request into unrelated product copy, backend, analytics, or global restyling without evidence that the requested flow …
- Very specific: line 96
Native Qt/QML custom control: verify actual Tab/focus-scope traversal, visible focus, documented keyboard activation, and Qt Quick …
- Distinctive: line 100
Native Qt/QML modal/dialog: verify initial focus, keyboard operation, modal containment when applicable, and focus return to the invoking …
- Distinctive: line 103
Native Qt/QML form/error: verify keyboard reachability and that error/status text is exposed through the native accessibility path.
- Distinctive: line 106
End-to-end retouch workflow: verify the real user path (for example, import, adjust, apply/save) in the running native app before calling …
- Very specific: line 110
Modal dialog: verify that opening moves focus into the dialog, Tab and Shift+Tab stay in its sequence, Escape closes it when the product …
- Distinctive: line 115
HTML/UXP form error: verify a detected invalid field is identified and its error is described in text.
- Very specific: line 124
Web motion: when motion changes, exercise the reduced-motion branch rather than assuming that a static source review proves it.
- Distinctive: line 128
Native motion: verify the product's documented OS/app motion accommodation if it exists.
- Very specific: line 142
Do not run or report test_catalog_refresh.py or test_relevance_evaluator.py as local PASS criteria: each imports an upstream-root script …
“Common” is measured: kept by at least 6 in 10 of the 49 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
UI Design
Use one product-surface loop. Preserve existing brand tokens and component conventions as the authority; do not replace them with a generic style guide.
This is the shared entry point for interface design. Supporting skills below are implementation details, not alternative design workflows. Read the shared contract once; a link back from a supporting skill is not a recursive reload.
Editable visual blocks
For an illustrated hero, promotional banner, product infographic, or a supplied visual to rebuild as editable UI, read Editable visual assets. Keep text, controls, data, and layout native to the product; generate only the imagery that the requested composition needs. A screenshot is a preview, not editable source or interaction proof. Ordinary forms and small CSS fixes skip this mode. Standalone posters and PowerPoint conversion are separate deliverables, not mandatory stages of UI work.
Select the smallest supporting skill set
- Read
../ui-ux-pro-max/SKILL.mdfor a new surface, substantial redesign, or a visual/UX decision that needs evidence. Use its local search scripts with a short, specific query. Treat a zero-result search as zero evidence; do not invent a recommendation. Do not persist a generated design-system file unless the task explicitly asks for it. - Read
../frontend-design/SKILL.mdwhile changing an HTML, browser, or UXP frontend. Keep the implementation native to the detected stack and existing component system. Do not apply its browser CSS/ARIA recipes to a native Qt/QML surface. - Read
../motion-framer/SKILL.mdonly when the changed surface is React/JS and the project already usesmotion/framer-motion, or the user has explicitly authorized adding it. Otherwise use the platform's native primitives; never add the package merely to animate a control. - For a running surface, use the inspection tool that matches the surface: browser screenshot, interaction, accessibility-tree, and visual-diff tools are for browser/UXP/HTML. For native Qt/QML, use available native desktop and accessibility inspection instead. Never require a browser tool to prove a native application.
- For native Qt/QML, keep existing Qt Quick Controls and theme components as the authority. When custom controls or focus behavior change, consult the Qt Quick input-focus and accessibility guidance; verify native accessible semantics and keyboard behavior rather than substituting ARIA or CSS rules.
Do not load all four by default. A focused CSS fix normally needs only the implementation guidance and a focused proof.
Work in one loop
- Inspect the target surface, real stack, existing tokens/components, changed user flow, supported widths, and any current visual evidence.
- State a compact design contract before editing: primary user action,
information hierarchy, reuse/new tokens, keyboard/contrast behavior, and
platform-appropriate sizing and motion behavior. Browser breakpoints and
prefers-reduced-motionare conditional, not universal desktop requirements. For a tiny local fix, keep this contract proportional to that fix. - Implement the visual and interaction changes together. Prefer semantic structure, stable component boundaries, and design tokens over one-off pixel overrides. Do not add animation that obscures feedback, delays an action, or ignores reduced-motion preferences.
- Verify the changed flow at the narrowest real boundary: project checks first, then live behavior appropriate to the target when runnable and the risk warrants it. Use browser/UXP screenshot and accessibility-tree evidence for web surfaces; use native Qt/QML keyboard and accessibility inspection for desktop surfaces. Check the affected platform sizing, keyboard path, visible focus, contrast, overflow, and motion behavior rather than declaring a page "polished" from source code alone. A source read or screenshot without an identified running build is never runtime PASS.
Decision boundaries
- Keep an established design system unless the task is explicitly a redesign.
- Keep UI Pro Max as retrieval evidence, not as an authority over the product's brand, accessibility contract, or existing UX research.
- Treat motion as progressive enhancement. The same action and information must remain clear with reduced motion and without animation support.
- Do not broaden a UI request into unrelated product copy, backend, analytics, or global restyling without evidence that the requested flow requires it.
Interaction-specific proof
Apply only the slice that the changed surface contains; this is not a demand to add a modal, custom validation, or extra animation to an otherwise simple UI. Browser-specific ARIA, CSS, and WCAG-pixel checks do not prove a native QML implementation; source inspection alone does not prove any runtime behavior.
- Native Qt/QML custom control: verify actual
Tab/focus-scope traversal, visible focus, documented keyboard activation, and Qt Quick accessible role/name/state with a native accessibility inspector where available. Keep a runtime receipt tied to the build identity. - Native Qt/QML modal/dialog: verify initial focus, keyboard operation, modal containment when applicable, and focus return to the invoking control after close.
- Native Qt/QML form/error: verify keyboard reachability and that error/status text is exposed through the native accessibility path; do not infer it from a property declaration.
- End-to-end retouch workflow: verify the real user path (for example, import, adjust, apply/save) in the running native app before calling the surface usable; code or a static screenshot is only source/visual evidence.
- Modal dialog: verify that opening moves focus into the dialog,
TabandShift+Tabstay in its sequence,Escapecloses it when the product supports dismissal, and closing restores focus to the invoking control (or a documented logical successor). Do not claimaria-modalunless the background is actually inert. WAI-ARIA APG modal-dialog pattern - HTML/UXP form error: verify a detected invalid field is identified and its error is described in text. An inline message, summary, alert, or native validation can be appropriate only when the real browser/assistive-technology behavior supports the chosen path; color alone is not an error description. WCAG 2.2 SC 3.3.1
- HTML/UXP dense pointer controls: for adjacent compact controls, assess the WCAG 2.5.8 boundary: a 24 by 24 CSS-pixel target or its spacing/equivalent/inline exception. Do not blindly enlarge an inline link or an essential dense control. WCAG 2.2 SC 2.5.8
- Web motion: when motion changes, exercise the reduced-motion branch rather
than assuming that a static source review proves it.
prefers-reduced-motioncommunicates the user's request to reduce, remove, or replace non-essential motion. MDN reference - Native motion: verify the product's documented OS/app motion accommodation if it exists. Do not invent a CSS media-query requirement for QML.
Local UI Pro Max validation boundary
The vendored ui-ux-pro-max folder is a distributed skill, not its complete
upstream repository. Its local acceptance checks are the data validation and a
real, scoped retrieval query, for example:
python skills/frontend/ui-ux-pro-max/scripts/validate_data.py
python skills/frontend/ui-ux-pro-max/scripts/search.py "keyboard focus modal" --domain uxDo not run or report test_catalog_refresh.py or test_relevance_evaluator.py
as local PASS criteria: each imports an upstream-root script absent from the
distributed skill. Preserve that boundary as NOT_RUN_UPSTREAM_DEPENDENCY, not
as a failure repaired with stubs or as a passing test.
Gotchas
- A fixed-canvas poster can look excellent while being unsuitable for a responsive interface. Do not import its absolute-pixel, single-file layout policy into a product.
- A selectable raster layer is not editable text, a working control, or a data-bound chart. Name the actual editability boundary.
Troubleshooting
- The agent found only
frontend-design: follow its shared-entry link here once, then load only the reference needed for the changed surface. - A copy change requires regenerating the hero: extract the copy into native text and keep the illustration separate; verify wrapping and contrast.
- The render is correct but an action cannot be used: inspect the running control, focus, and hit area. More screenshot polishing does not fix it.
Other versions59 projects use a version of this.
Word for word · 1
- AnastasiyaW/codex-claude-code-config★ 213100% same
Edited · 0
None found.
Wrote their own version · 58
- epicenter-md/epicenter★ 4.8ksame name
- EpicenterHQ/epicenter★ 4.8ksame name
- yunshu0909/yunshu_skillshub★ 760same name
- Mindrally/skills★ 275same name
- mblode/agent-skills★ 177same name
- JayZeeDesign/awesome-claude-skills★ 169same name
- frappe/skills★ 162same name
- TencentCloudBase/skills★ 80same name
- getsentry/abacus★ 77same name
- uizze/uizze★ 77same name
48 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.
Used alongsideProjects with it also use graphic-design, typography and 9 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.
travisjneumanSkillgraphic-design
Professional graphic design principles for digital and print media. Use when creating visual designs, choosing color palettes, typography, layouts, or providing design feedback.
Also in 3 projects that use this
ArnavPuriSkilltypography
Typography for web and UI: modular type scales, font pairing, heading hierarchy, readable body text, fluid clamp() sizing, variable fonts, font loading, and text effects. Use when the task is about fonts or text styling. Trigger phrases: "typography", "font pairing", "type scale", "heading sizes", "responsive text", "font combination", "text style", "variable fonts", "fluid typography", "google fonts", "line height", "font loading". For a full token system (spacing, color, components) use design-system, which consumes the scale defined here.
Also in 3 projects that use this
MicrosoftSkilltypescript
TypeScript best practices for this monorepo. Use this skill when: (1) writing new TypeScript code, (2) reviewing TypeScript code, (3) working with API responses, fetch calls, or JSON parsing, (4) adding or modifying types and interfaces.
Also in 7 projects that use this
OpenAISkillauth
Authentication integration guidance — Clerk (native Vercel Marketplace), Descope, and Auth0 setup for Next.js applications. Covers middleware auth patterns, sign-in/sign-up flows, and Marketplace provisioning. Use when implementing user authentication.
Also in 3 projects that use this
VercelSkillreact
React renderer for json-render that turns JSON specs into React components. Use when working with @json-render/react, building React UIs from JSON, creating component catalogs, or rendering AI-generated specs.
Also in 4 projects that use this
3stoneBrotherSkillcode-audit
Professional code security audit skill covering 55+ vulnerability types. Enhanced with WooYun 88,636 real-world vulnerability cases (2010-2016). This skill should be used when performing security audits, vulnerability scanning, penetration testing preparation, or code review for security issues. Supports 9 languages: Java, Python, Go, PHP, JavaScript/Node.js, C/C++, .NET/C#, Ruby, Rust. Includes 143 mandatory detection items across all languages with language-specific checklists. Covers SQL injection, XSS, RCE, deserialization, SSRF, JNDI injection, JDBC protocol injection, authentication bypass, business logic flaws, race conditions, and modern security domains (LLM, Serverless, Android). WooYun integration adds: statistical-driven parameter priority, bypass techniques library, logic vulnerability patterns, and real-case references. v1.0: Initial public release with Docker deployment verification framework.
Also in 3 projects that use this
In this repository60 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- agent-harness-designSkill
- agents-best-practicesSkill
- diffusion-engineeringSkill
- flux2-klein-promptingSkill
- flux2-lora-trainingSkill
- forensic-prompt-compilerSkill
- ml-research-labSkill
- notebooklm-grounded-researchSkill
- vlm-segmentationSkill
- feature-newSkill
- harness-designSkill
- layer-newSkill
- plan-swarm-reviewSkill
- pixel-art-storyboardSkill
- pixel-art-studioSkill
- architecture-firstSkill
- architecture-qualitySkill
- bug-reproducerSkill
- code-complexitySkill
- control-cliSkill
- control-uiSkill
- deep-reviewSkill
- deslopSkill
- distill-feedbackSkill
- harness-feedbackSkill
- proof-verifySkill
- refactoring-safelySkill
- repo-mapSkill
- root-cause-deliverySkill
- system-and-data-designSkill
- testing-strategySkill
- thermo-nuclear-code-quality-reviewSkill
- verify-thisSkill
- workflow-orchestrationSkill
- frontend-designSkill
- motion-framerSkill
- ui-ux-pro-maxSkill
- ios-developmentSkill
- lean-codeSkill
- cross-harness-continuationSkill
- deepseek-provider-contractSkill
- desktop-sessions-discoverySkill
- epistemic-challengeSkill
- gemini-delegateSkill
- harness-auditSkill
- observability-monitoringSkill
- remote-compute-opsSkill
- plan-to-ticketsSkill
- product-meaning-extractorSkill
- remotion-production-guideSkill
- script-evaluatorSkill
- video-narrative-arcSkill
- video-post-productionSkill
- article-structure-reviewSkill
- cavemanSkill
- humanize-englishSkill
- humanize-russianSkill
- owner-voiceSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
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.

