Use this skill when the user needs production-ready frontend design and implementation: designing or refining components, defining a visual system, building responsive layouts, implementing polished interactions, improving accessibility, reducing generic AI-looking UI, or preparing distinctive UI changes for review, merge, or handoff while preserving upstream workflow, copied support files, and provenance.
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Use this skill when the user needs production-ready frontend design and implementation: designing or refining components, defining a visual system, building responsive layouts, implementing polished interactions, improving accessibility, reducing generic AI-looking UI, or preparing distinctive UI changes for review, merge, or handoff while preserving upstream workflow, copied support files, and provenance.
Use this skill when the task requires both design judgment and implementation discipline. The goal is not to generate a generic layout, but to produce a distinctive interface that is semantically sound, accessible, responsive, maintainable, and ready for real review or handoff.
Preserve the upstream identity and provenance. Keep copied support files intact unless the task explicitly requires changes, and do not hide the source workflow when handing work back.
When to Use
Use this skill when the user needs to:
design or refine a UI with a clear visual point of view
implement production-grade frontend components, screens, or flows
replace generic or template-like styling with intentional systemized design
improve accessibility, semantics, interaction quality, or responsive behavior
encode visual decisions into reusable tokens, layers, and component rules
prepare frontend work for review, merge, or handoff with explicit quality gates
Do not use this skill as the primary workflow when the task is mainly:
backend-only work
copywriting without UI implementation
logo or brand exploration with no frontend output
quick throwaway mockups where production constraints do not matter
Workflow
Clarify the product and interaction constraints
Identify the page, component, or flow being changed.
Capture target users, key actions, states, and device contexts.
Note hard constraints: framework, design system, browser support, performance budget, and existing tokens.
Ask what must stay stable versus what can change visually.
Audit the current interface before redesigning
Inspect structure, semantics, and existing components.
Record what already works so redesign does not erase useful behavior.
Define the design direction in system terms
Describe the intended aesthetic using concrete implementation levers: spacing rhythm, type scale, density, contrast model, border treatment, motion style, elevation, and color roles.
Convert one-off visual ideas into reusable tokens or component rules.
Prefer a small number of strong, repeatable decisions over decorative noise.
Build semantic structure before ornament
Start with correct landmarks, headings, lists, buttons, links, forms, and status messaging.
Ensure the interface still makes sense with styles reduced.
Use native elements first; add ARIA only where native HTML cannot express the needed behavior.
Implement responsive layout with modern CSS primitives
Use layout systems that adapt by container or content, not only by viewport.
Prefer flex, grid, intrinsic sizing, clamp(), and container queries where supported by project constraints.
Use cascade layers and predictable selectors to avoid style collisions.
Design interactions and states explicitly
Implement hover, focus, active, disabled, loading, empty, error, and success states.
Make focus appearance obvious and keyboard traversal coherent.
Keep motion subtle, purposeful, and respectful of reduced-motion preferences.
Run production gates before handoff
Check keyboard access, visible focus, contrast, target size, responsive breakpoints or containers, content reflow, and state coverage.
Review likely performance regressions, especially LCP and CLS risks from fonts, images, and late-loading UI.
Confirm the final output preserves provenance and any upstream support files required by the source workflow.
Output Expectations
When executing this skill, the final work should usually include:
the implemented UI change or component code
a brief explanation of the design direction
any new or modified tokens, variables, or component conventions
accessibility notes for semantics, keyboard behavior, and state handling
responsive notes for layout adaptation
known tradeoffs, constraints, or follow-up items
Best Practices
Do:
make the structure correct before styling it
use native controls for buttons, links, inputs, dialogs, and disclosure patterns whenever possible
encode repeated visual choices as tokens, utilities, or component-level rules
use container-aware layout logic when components live in multiple contexts
verify empty, loading, error, and disabled states instead of styling only the default happy path
keep distinctive design grounded in hierarchy, rhythm, contrast, and interaction clarity
check whether fonts, media, shadows, animation, or script-driven layout changes could hurt performance
Do not:
fake buttons or links with non-semantic elements unless there is a justified accessibility-safe pattern
rely on placeholder text, color alone, or motion alone to communicate critical state
add visual complexity that introduces layout shift, weakens readability, or hides task priority
hard-code one-off spacing, colors, or breakpoints when the decision should become a reusable token
treat accessibility as a post-processing step after the interface is already shaped
assume viewport breakpoints alone are enough for reusable components
Troubleshooting
Symptoms: The UI looks polished visually but keyboard navigation fails or focus disappears.
Solution: Replace fake controls with native elements where possible, restore visible focus styles, verify tab order, and ensure interactive elements are reachable and operable without a pointer.
Symptoms: The design feels distinctive in one screen size but breaks when embedded in a narrower panel or card.
Solution: Rework the component around intrinsic sizing, flex/grid constraints, and container-based behavior. Remove assumptions tied only to full-page viewport width.
Symptoms: Styles become hard to override or small changes trigger regressions elsewhere.
Solution: Move repeated choices into tokens, reduce selector specificity, and separate reset/base/component/utility concerns with cascade layers or equivalent project structure.
Symptoms: Rich visuals cause cumulative layout shift or slow first render.
Solution: Reserve space for media, avoid late dimension changes, review font loading strategy, reduce unnecessary animation on load, and inspect whether decorative assets are affecting LCP.
Symptoms: The interface passes visually but still feels generic.
Solution: Tighten a few core system choices instead of adding more decoration: type hierarchy, spacing rhythm, corner and border language, contrast strategy, motion behavior, and component density.
Examples
Example 1: Replace a generic card grid with a reusable, production-ready system
Input
Create a pricing section with three cards. Make it modern.
Better execution
Design and implement a pricing section with three plans using semantic headings, list-based feature groups, a clear primary plan, visible keyboard focus, and responsive behavior that works both full-width and inside a narrower marketing shell. Define spacing, radius, border, and color decisions as reusable tokens instead of one-off values. Include hover, focus, disabled, and reduced-motion-safe interaction states.
Expected outcome
one plan has clear emphasis without breaking contrast or readability
cards stack or reflow cleanly in narrow containers
CTA controls are real buttons or links with visible focus
spacing and color values are reusable in nearby sections
Example 2: Improve an existing component instead of repainting it
Input
Make this dashboard sidebar look less boring.
Better execution
Audit the existing dashboard sidebar and improve it without changing information architecture. Preserve working navigation behavior, then refine hierarchy, spacing rhythm, icon alignment, active-state treatment, and focus appearance. Remove generic framework defaults, convert repeated style decisions into tokens, and verify the sidebar remains usable at narrower container widths.
Expected outcome
navigation remains structurally stable
active, hover, and focus states are differentiated clearly
the sidebar looks intentional rather than template-derived
no regressions in keyboard use or responsive containment
Open examples/worked-example.md for a concrete before/after implementation pattern.
Additional Resources
references/domain-notes.md - Open this during implementation or review for compact production guidance on semantics, accessibility, responsive CSS, tokens, and performance checks.
examples/worked-example.md - Open this when you want a concrete transformation from a generic UI request into a production-grade implementation plan and acceptance criteria.
Related Skills
No local related skills were provided in the source context. Do not invent substitutes.
