agentleFS
kozz36kozz36/frontend-designer-skill/skills/frontend-designer-lite/SKILL.md

frontend-designer-lite

Skill9 starsChanged 3 months agoApache-2.0View on GitHub

Trigger: frontend design, CSS architecture, tokens, responsive UI, accessibility, motion. Produce production UI design-system decisions without local references.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 12 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈12

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this standalone skill for frontend design-system decisions when an agent must choose, review, or document production-ready technical …

    line 12
  2. Distinctive:

    Do not use it for generic explanation, copy editing, or one-off code changes that do not affect architecture or reusable implementation …

    line 18
  3. Distinctive:

    Use semantic HTML and native controls before ARIA substitutes.

    line 23
  4. Distinctive:

    Keep components reusable and responsive.

    line 24
  5. Distinctive:

    Keep the answer decision-first.

    line 25
  6. Distinctive:

    Treat security requirements as conditional product and repository inputs.

    line 26
  7. Distinctive:

    Tooling: run only repository-provided validation with an available dependency, target, permission, and expected evidence.

    line 39
  8. Distinctive:

    Do not add browser, API, accessibility-law, security, adoption, or version claims without checking a current authoritative source.

    line 43
  9. Distinctive:

    Use the relevant authoritative source: W3C CSS specifications for syntax, MDN for browser-feature behavior, WCAG for applicable …

    line 46
  10. Distinctive:

    Stop for missing required evidence.

    line 51
  11. Distinctive:

    Verify any time-sensitive version, browser, API, accessibility-law, security, adoption, or legal claim against current authoritative …

    line 53
  12. Distinctive:

    Run available repository-native validation and report unavailable validation as evidence still needed.

    line 54

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Activation Contract

Use this standalone skill for frontend design-system decisions when an agent must choose, review, or document production-ready technical direction.

  • Designing or reviewing UI systems, component anatomy, tokens, layout, or visual interaction.
  • Choosing CSS architecture, responsive behavior, accessibility constraints, or motion rules.
  • Converting repository visual direction into implementation-ready frontend guidance.

Do not use it for generic explanation, copy editing, or one-off code changes that do not affect architecture or reusable implementation patterns.

Non-Negotiable Runtime Rules

  • Preserve repository token and component-state contracts; do not hardcode isolated visual decisions.
  • Use semantic HTML and native controls before ARIA substitutes. Define native keyboard behavior, visible focus, measurable contrast, touch/pointer criteria, and reduced-motion behavior for every interactive UI.
  • Keep components reusable and responsive; reserve media and font geometry and do not accept performance regressions without compatible evidence.
  • Keep the answer decision-first. Do not add version, browser, API, accessibility-law, security, adoption, or legal-applicability claims without current authoritative verification.
  • Treat security requirements as conditional product and repository inputs; do not infer controls, policy, telemetry, or external-tool permission when evidence is absent.

Required Evidence Gates

Gather the following before selecting a pattern. If a required input is unavailable, stop and report the missing evidence rather than guessing or prescribing a universal stack.

  • Repository architecture and migration: inspect existing CSS architecture, naming, linting, build pipeline, team conventions, and accepted migration cost. Preserve the established paradigm; introduce cascade layers, CSS Modules, CSS-in-JS, utilities, or naming conventions only with repository support and an accepted migration plan.
  • Design source: inspect Design.md, design.md, token sources, a design-system package, design-tool export, or another approved source when present. Treat its case-sensitive path and token, semantic, component, and state contracts as local visual truth; reconcile conflicts with code and team conventions. Do not require a filename, rename a source, or assume a design tool.
  • Browser and fallback policy: inspect the supported browser matrix before using CSS nesting, container queries, :has(), light-dark(), scroll-driven animation, WAAPI, observers, or related APIs. Define a compatible fallback when required.
  • Tokens, color, and themes: preserve primitive, semantic, and component tiers or the local equivalent, including loading, disabled, error, success, and selected states. Use OKLCH-only for newly specified tokens only when the approved token contract and browser/build matrix select it; otherwise retain the compatible format. Measure every foreground/background and focus pair. Use dark-mode-first only when product requirements and the approved design source require it; otherwise preserve the existing semantic-token and theme mechanism.
  • Layout and components: select component-local container queries only when supported and valuable; use viewport queries for document-level layout, navigation, or the established responsive system. Choose mobile-first, desktop-first, grid, flex, named areas, or composition from information architecture, content, component ownership, and viewport evidence. Use semantic structure: accessible names for regions when needed, native <dialog> when it fits policy, native table semantics for tabular data, explicit form labels/errors/order/states, and <nav> with list semantics where applicable. Reserve media geometry with aspect-ratio, dimensions, or an equivalent contract when it reduces measured layout shift.
  • Accessibility and motion: preserve semantic HTML, labels, names, roles, states, keyboard order, persistent visible :focus-visible, unobscured focus, measurable contrast, and applicable touch-target criteria. Add ARIA only when native semantics cannot express the contract. Honor prefers-reduced-motion with equivalent low-motion or static feedback. Use the smallest motion mechanism and gate WAAPI, scroll-driven CSS, observers, and listeners on browser support, motion policy, and measured main-thread cost. Apply jurisdiction-specific obligations only when product scope establishes them.
  • Visual treatment and performance: use bento-like grids, glass, gradients, bold type, dark-first presentation, or other trends only when the approved design source, hierarchy, contrast, focus, affordance, responsive behavior, browser matrix, and performance evidence support them. Use content-visibility or contain only after checking rendering and accessibility behavior and measuring the target. Use will-change transiently and only when profiling justifies it. Choose image/font formats, dimensions, loading, fallbacks, and metrics from the delivery pipeline, browser matrix, and layout-shift evidence.
  • Tooling: run only repository-provided validation with an available dependency, target, permission, and expected evidence. Prefer documented package-manager scripts. Do not introduce implicit installs, external audits, encoding commands, or design-tool commands when prerequisites are absent.

Live Verification Rules

  • Do not add browser, API, accessibility-law, security, adoption, or version claims without checking a current authoritative source.
  • A URL is a starting point, not proof that a feature fits the repository browser matrix, product scope, build pipeline, or fallback policy.
  • Record verification date, source URL, scope, and confirmed behavior for every time-sensitive release or decision claim.
  • Use the relevant authoritative source: W3C CSS specifications for syntax, MDN for browser-feature behavior, WCAG for applicable accessibility criteria, and the official regulator for jurisdictional applicability. Check OKLCH and delivery choices against the approved pipeline and browser policy rather than claiming universal support.

Execution Order

  1. Identify repository architecture, design source, browser matrix, accessibility and legal scope, performance evidence, team conventions, tooling availability, migration cost, and security constraints.
  2. Stop for missing required evidence. Otherwise preserve token, state, semantic HTML, focus, contrast, reduced-motion, and media-geometry contracts before visual treatment.
  3. Select the smallest compatible CSS paradigm, layout, component, token, theme, media, and motion approach. Record required fallbacks, migration steps, and rejected alternatives with their concrete tradeoffs.
  4. Verify any time-sensitive version, browser, API, accessibility-law, security, adoption, or legal claim against current authoritative evidence before using it.
  5. Run available repository-native validation and report unavailable validation as evidence still needed; do not invent a successful command or result.

Output Contract

Return:

  • The recommended pattern, repository and source evidence, and why it fits the established contracts.
  • Rejected alternatives with concrete tradeoffs and migration cost where relevant.
  • Accessibility, browser, performance, security, and runtime risks, with mitigations and compatible fallbacks.
  • A fallback or migration plan, including conditions that require it.
  • Validation output: commands actually available and their results, plus missing evidence or verification still required before adoption.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.