Cómo usar
- Copia el comando y revisa el paquete antes de instalarlo en tu entorno de agente.
- Consulta las herramientas, credenciales y entornos necesarios en el documento original.
- Describe las entradas y el resultado deseado; contrasta la salida con las fuentes.
npx skills add https://github.com/jaimejunr/claude-code-frontend-skills --skill frontend-stackAntes de empezar
- La instalación no incluye suscripción al modelo, cuota API ni permisos sobre tus archivos.
- Las estrellas e instalaciones reflejan actividad, no una prueba independiente de calidad o seguridad.
Contenido del repositorio de origen con títulos, ejemplos, código, tablas, enlaces e imágenes preservados.
Frontend Stack router
Several design packs are installed side by side. Each is good alone and each assumes it is the only one in the room. This skill decides who leads in each phase of the work, so the agent loads one lead and at most two supporting skills at a time instead of five competing rulebooks.
frontend-design:frontend-design is the base. Load it first on every frontend task, whoever leads, and keep it loaded. It sits outside the lead/support count in step 7. If it is not in the session's skill list, say so once and go on.
1. Read the project first
Before picking anything, look for what the project already decided:
PRODUCT.md/DESIGN.mdat the repo root (Impeccable's context files)design-system/*/MASTER.md(UI UX Pro Max--persistoutput)docs/brand-guidelines.md,assets/design-tokens.*(UI UX Pro Maxbrand
and design-system)
- an existing design system: tokens,
tailwind.config.*,components/ui/, a
component library in package.json
- a brand guide the user pointed at
If any of these exist, follow them by default (rules 1 and 2 of the precedence ladder). A personality, identity or "looks AI-made" complaint authorizes a replacement visual world within the complaint's scope, even with DESIGN.md.
Keep one source of truth. If two of these files exist and disagree (palette, fonts, spacing), ask the user which one wins before writing code. When the project has none, use the identity flow below for a new brand, or Impeccable document to capture an established world. DESIGN.md is the source of truth; do not also run --persist or generate docs/brand-guidelines.md unless the user asks, and if they do, keep the values identical to DESIGN.md.
2. Classify the request
Answer three questions before picking skills:
- Job. Build something new; change something that exists (redesign,
restyle, migrate to a library); review only (critique, audit, polish); one craft (motion, copy, color, type, layout, tokens); or assets (brand, logo, banners, slides).
- Surface. Marketing (landing, portfolio, site); product (dashboard, chat,
forms, settings, tables, flows); a single component; mobile web; native app.
- Stack signals from step 1: a component library in
package.json, shadcn
(components.json, components/ui/), Tailwind, React Native/Expo, Swift.
Complaints about looks ("ugly", "generic", "looks AI-made", "slop", "cluttered") mean the job is change, not review: the user wants a new direction, not only a list of problems. An explicit review-only request stays read-only, even when it names an identity failure.
Vague complaint: ask before loading
A complaint is vague when it names no concrete problem, no "what" is wrong ("your site is still bad", "this screen is ugly", "the UX is not good", "improve the identity"). A known target ("this screen", a named page) does not make it concrete: it only skips the "where" question. A broad word (UX, usability, experience, design, identity, navigation) is still vague, but it picks a base path. Then do the basics only:
- Load
frontend-design:frontend-designand nothing else. Read the project
(step 1) and the conversation: "still" means an earlier round failed, so find what that round changed.
- Pick the base path in base-paths.md from the
complaint's words: UX, navigation, reading and hierarchy, flows and forms, starting or improving an identity, looks, copy, mobile. No area named: the first question asks which one.
- If a browser or screenshot tool is at hand and the app runs, look at the
current render of the page the user most likely means, to ground the questions. What you saw in the code or the render only shapes the options; do not list problems yet.
- Ask the path's questions in one message, with the question tool when the
session has one: at most four, each with options. Skip what the context answers ("this screen" with a screenshot, an open URL or the last page changed). If an earlier round failed, ask what in it missed.
- Wait for the answers. Route each answer as its path says; several answers
become phases (step 3).
Before the answers: no critique, list of problems found, direction, prototype, recommendation or other skill load. Describe the project in one neutral line (what pages exist), not what is wrong with them. If the user declines to answer or says "you decide", follow the path's fallback, or the shared one in base-paths.md; on the looks path or with no area named, treat it as a looks complaint about the main page and run the gate in step 3.
3. Work in phases
Bigger tasks run in phases, in this order. Load each phase's skills when the phase starts, not all up front. Skip a phase that the task does not need.
| Phase | When | Lead | Support |
|---|---|---|---|
| Diagnose | the job is change | impeccable:impeccable (critique) | impeccable:impeccable (audit) for a11y/perf debt |
| Direction | new work, or any complaint about looks | identity flow below when the product has no brand or its identity is being replaced; otherwise product: frontend-design:frontend-design, marketing: taste:design-taste-frontend | reference research before directions; ui-ux-pro-max:ui-ux-pro-max loaded at the start for UX and stack rules, its styles and presets only after the concept |
| Build | always | the project's component library (its docs, components, tokens, MCP server if any) | new marketing pages: impeccable:impeccable via reference/new-work.md; ui-ux-pro-max:ui-styling when shadcn/Tailwind |
| Motion | the change moves | emil-design-eng:animate | emil-design-eng:animation-vocabulary |
| Review | anything user-facing | see step 8 | - |
At the start, tell the user the phase plan with the skills of each phase, so they see the whole route even though only the first phase is loaded.
For a new landing with no brand, or a complaint about personality, identity, generic output or "looks AI-made", use the identity flow below first.
For any other complaint about looks, Diagnose and Direction run together: load impeccable:impeccable (critique), the Direction lead and ui-ux-pro-max:ui-ux-pro-max at the start, always, even when the request names no stack or style. A critique alone is a list of problems, not the new direction the user asked for. This is a gate: call the Skill tool for each of them, plus frontend-design:frontend-design, before the first diagnosis, direction, prototype or recommendation, and name them in the phase plan. Directions proposed without them loaded are void: load them and redo the work. This overrides "load each phase's skills when the phase starts" above. A vague complaint (step 2) runs its questions first; the gate applies once the answers place it here. Research references before proposing directions (see identity-flow.md, step 2); reference-gallery.md seeds the board with inspected modern sites, grouped by the problem each one solves. UI UX Pro Max supplies UX and stack rules; category styles and presets are finishing material after the concept, never its source. When the project has a design system, use UI UX Pro Max for UX and stack rules only, not to replace the palette or fonts (rule 2 of the ladder).
Write what Diagnose and Direction decided (in DESIGN.md or the plan) before Build, so later phases do not need the earlier skills loaded.
When the product has no identity
Read references/identity-flow.md. It owns the full route: product context and naming/positioning → reference board → concept devices and Impeccable concept seed → 2-3 directions → prototype and user pick → brand → DESIGN.md and tokens → landing → screenshot finish gate.
The user picks the direction. Preserve their approved concept and praised or fixed visual choices. Presets finish a direction; they never generate it.
4. Pick the skills for the task
| The task is... | Lead | Support |
|---|---|---|
| New landing for a product with no brand | the identity flow in step 3, before the landing | references/identity-flow.md; brand → system → Build |
| Landing page for a developer tool / API / AI product | taste:design-taste-frontend, or the identity flow first when there is no brand | devtool-landing.md + landing-anatomy.md; impeccable:impeccable new-work in Build |
| New landing page, portfolio, marketing site with a brand | taste:design-taste-frontend | frontend-design:frontend-design for direction; landing-anatomy.md; impeccable:impeccable new-work in Build |
| New marketing page where the user names Hallmark, one of its themes, or asks for a themed or custom page | hallmark:hallmark (leads instead of Taste, never with it) | with a brand or DESIGN.md it reads that as the locked system and deals no theme; landing-anatomy.md |
| New product UI: dashboard, settings, tables, multi-step flows | frontend-design:frontend-design | ui-ux-pro-max:ui-ux-pro-max (stack + UX guidelines), impeccable:impeccable for harden at the end |
| Redesign or restyle an existing product app | phases: Diagnose → Direction (frontend-design:frontend-design) → Build → Review | ui-ux-pro-max:ui-styling in Build when shadcn/Tailwind |
| Redesign an existing marketing site | taste:design-taste-frontend (it audits the live page before changing it), or hallmark:hallmark (redesign <path>) when the user names Hallmark | impeccable:impeccable (critique) first |
| No personality, no product identity, "looks rushed" | the identity flow in step 3 | - |
| "Looks generic / AI-made / slop" | the identity flow in step 3 | frontend-design:frontend-design anti-template checks during Review |
| Adopt or migrate to a component library or design system ("use our kit", "move to shadcn") | the library itself: its docs, components, tokens | ui-ux-pro-max:ui-styling for Tailwind/shadcn mapping, impeccable:impeccable (extract) to pull tokens |
| Chat or AI assistant UI | frontend-design:frontend-design + the library's chat components if it has them | impeccable:impeccable (harden) for empty, error, loading and streaming states; emil-design-eng:animate for streaming motion |
| Forms, onboarding, first-run | impeccable:impeccable (onboard, harden) | frontend-design:frontend-design |
| Empty, error, loading states, edge cases | impeccable:impeccable (harden) | - |
| Only critique or audit, no rebuild asked | impeccable:impeccable (critique then audit); hallmark:hallmark (audit <path>) when the user names Hallmark | none: the user asked for a report, not a new direction |
| "Plan it before coding", discovery, UX flow | impeccable:impeccable (shape) | - |
| Broad area, no detail ("improve the UX", "the navigation is bad", "work on the identity") | its base path in base-paths.md: questions first | the skills the answers route to |
| Vague brief, no direction yet | frontend-stack:design-director agent, then route its result here | - |
| Review a page, component or PR diff | frontend-stack:ui-reviewer agent | - |
| Set up design context for a repo | impeccable:impeccable (init, then document) | - |
| Polish before shipping | impeccable:impeccable (polish) | emil-design-eng:review-animations for interaction details |
| Accessibility, performance, responsive checks | impeccable:impeccable (audit, optimize, adapt) | - |
| Too loud, or too bland | impeccable:impeccable (quieter / bolder) | - |
| Only typography / only color / only layout | impeccable:impeccable (typeset / colorize / layout) | ui-ux-pro-max:ui-ux-pro-max for font pairings or palettes |
| Choose a style, palette, font pairing, chart type | ui-ux-pro-max:ui-ux-pro-max | - |
| Charts and data visualization | ui-ux-pro-max:ui-ux-pro-max (chart guidance) | emil-design-eng:pick-ui-library for the chart library |
| Dark mode, theming | ui-ux-pro-max:design-system | impeccable:impeccable (colorize) |
| Design tokens / design system architecture | ui-ux-pro-max:design-system | impeccable:impeccable (extract) |
| shadcn/ui + Tailwind component work | ui-ux-pro-max:ui-styling | emil-design-eng:pick-ui-library |
| Screenshot or mockup image to code | taste-imagegen:image-to-code (optional plugin) | frontend-design:frontend-design |
| Build an animation | emil-design-eng:animate | emil-design-eng:animation-vocabulary when the user describes motion vaguely |
| Review or fix existing motion | emil-design-eng:review-animations / improve-animations | - |
| "Where should this animate?" | emil-design-eng:find-animation-opportunities | - |
| Gesture, spring, Apple-feel UI | emil-design-eng:apple-design | - |
| Web app should feel native on phones | emil-design-eng:mobile-native | impeccable:impeccable (adapt) |
| Try several versions of a component | emil-design-eng:prototype | - |
| Which library for X (charts, OTP, virtual lists...) | emil-design-eng:pick-ui-library | - |
| Toasts with Sonner | emil-design-eng:ask-sonner | - |
| UX copy, errors, microcopy, empty-state text | ux-research:ux-copy | ui-ux-pro-max:brand voice when the project has one; impeccable:impeccable (clarify) to fix copy in an existing screen |
| User research plan, interview guide, usability test | ux-research:user-research | - |
| Interviews, usability notes, NPS or tickets turned into UX decisions | ux-research:research-synthesis | - |
| Brand voice, visual identity, brand guidelines | ui-ux-pro-max:brand | ui-ux-pro-max:design-system for tokens |
| Logo, icons, corporate identity (CIP), social photos | ui-ux-pro-max:design | - |
| Banners, ads, social images, website hero art | ui-ux-pro-max:banner-design | - |
| Slides, pitch deck | ui-ux-pro-max:slides | - |
| Brand-kit board as a generated image | taste-imagegen:brandkit (optional plugin, needs image generation) | - |
| Explicit look: minimalist, brutalist, "expensive agency" | taste-classics:minimalist-ui / industrial-brutalist-ui / high-end-visual-design (optional plugin) | taste:design-taste-frontend for the inspector and fix loop |
| Google Stitch DESIGN.md | taste-classics:stitch-design-taste (optional plugin) | - |
| React Native / Expo motion, Swift | emil-native:animate-expo / write-swift (optional plugin) | - |
| Mobile app screens designed with Sleek, then built in React Native or HTML | design-mobile-apps:design-mobile-apps (optional plugin; needs SLEEK_API_KEY, sends the brief to sleek.design, paid beyond a trial) only when the user asks for Sleek | emil-native:animate-expo for motion after the build |
| Video that goes with a UI: product demo, hero or explainer clip, changelog, captions, animated overlay | hyperframes:hyperframes (optional plugin; needs Node 22 and the hyperframes CLI) | hyperframes:media-use for music, voice and images; emil-design-eng:animate stays the owner of in-page UI motion |
| Mockup image first, then code | taste-imagegen:imagegen-frontend-web / imagegen-frontend-mobile (optional plugin, needs image generation) | - |
| Video written as React components (Remotion) | remotion-best-practices when it is in the session's skill list (Remotion's own skills, not vendored: npx skills add remotion-dev/skills; Remotion needs a paid company license for for-profit organizations above 3 people) | hyperframes:hyperframes if the user prefers plain HTML |
A request that mixes jobs ("redesign it and add animations") is several rows: run them as phases in the order of step 3.
5. Cases the table does not cover
New skills arrive with every upstream sync, and requests will not always match a row.
- No row fits: use the job and surface from step 2, give the lead to the
domain owner (rule 5 of the ladder), then check the session's skill list for a skill whose description fits better.
- A named skill is not in the session's skill list: skip it and use the
next row's lead or frontend-design:frontend-design. Never guess a name.
- Not a UI question: product strategy, positioning, what to build, specs,
roadmap, metrics, backend architecture or running the business belong to the companion stack. If indie-dev-stack:indie-dev-stack is in the session's skill list, hand the request to it. Otherwise do what you can and tell the user once per session that the companion marketplace covers it: /plugin marketplace add JaimeJunr/claude-code-indie-dev-stack.
- A skill not listed here fits better: a row still wins unless the user
names the other skill. Use it as support and tell the user this router has no row for it yet.
6. Precedence ladder
When two loaded skills disagree, the higher rule wins:
- The user's explicit request and brand. "Use Inter", "purple is our
brand", "keep it like Linear" beat every skill ban.
- The project's existing system.
DESIGN.md, tokens, the component
library already in use. Extend it; do not swap libraries or mix systems. A complaint about personality, identity or "looks AI-made" authorizes replacing the visual world for what the complaint covers, even if DESIGN.md exists. Keep only the visual choices the user explicitly praised or fixed in that scope. Preserve product truth, working behavior, accessibility and the framework/library; leave unrelated surfaces alone.
- Accessibility and correctness. Contrast, focus states, reduced motion,
keyboard access, no layout shift. Impeccable audit is the referee.
- Scope of the lead skill. A skill's own "not for X" line is binding.
design-taste-frontend now also covers app screens, but the stack still leads product UI with frontend-design unless the user asks for Taste.
- Domain owner. Motion questions go to Emil's skills, style/palette data to
UI UX Pro Max, critique/audit/polish to Impeccable, aesthetic direction to frontend-design or taste.
- The lead skill's taste defaults.
frontend-design's anti-template
rules outrank any other lead's taste defaults.
The rules that actually collide between packs, and how each is settled, are in references/conflicts.md. Read it when a supporting skill contradicts the lead.
7. Load budget
- Per phase: the base
frontend-design, one lead and at most two supporting
skills. Big skills (design-taste-frontend, ui-ux-pro-max, impeccable) are long; loading all of them at once burns context and produces averaged, generic output.
- Pick support from other packs than the lead when they own part of the
subject (rule 5 of the ladder): one pack alone loses the knowledge only the others carry, even when it has a command for everything.
- Say in one line which skills each phase loaded, so the user can correct the
routing early.
- Aesthetic presets (
taste-classics:minimalist-ui,industrial-brutalist-ui,
high-end-visual-design) never stack: at most one, as finishing material after the concept, or as the user's explicit choice.
- One marketing-page lead per task:
taste:design-taste-frontendor
hallmark:hallmark, never both. Hallmark's Floor binds only pages it leads.
- The brand, logo, banner and slides skills above overlap (
ui-ux-pro-max:design
and design-system also cover slides and branding). Load only the row's lead.
- Prefer the specific Emil skill over
emil-design-eng:emil-design-engas
support: without a direct question the umbrella replies with a stock greeting.
8. Finish the same way every time
For anything user-facing, end with a review pass before calling it done:
- Capture and open desktop + mobile screenshots with the session's available
browser/screenshot tool. Compare the render with the reference board and chosen concept devices. Without tooling, report the missing visual evidence; do not claim the screenshot gate passed.
impeccable:impeccableaudit(or delegate to theui-revieweragent).
For new marketing pages, Build follows Impeccable's reference/new-work.md and hands captures + direction contract to impeccable-finish-reviewer (plugins/impeccable/agents/impeccable-finish-reviewer.md in this repo). Follow its recapture / rebuild / fix / ship disposition; use its documented fallback only when agents are unavailable. A page hallmark:hallmark or taste:design-taste-frontend led ends with that skill's own inspector and fix loop in place of that screenshot gate; still run audit for a11y.
- If the change has motion,
emil-design-eng:review-animations. - Fix the findings and recapture affected views. For personality/identity/
AI-look requests, a missing signature or a render interchangeable with a named competitor can block completion. Report evidence, verdict and gaps.