Ux Skill

@Laith0003

a day ago

A deterministic, offline design-intelligence engine for AI coding tools. Replaces improvised UI generation with structured constraints plus a rule-based anti-AI-slop linter. Ships as a Claude Code plugin, a Python package, an MCP server (18 tools), and a 17-IDE installer.

概览工具评论

概览

English · العربية · 简体中文 · 繁體中文 · 日本語 · 한국어 · हिन्दी · Bahasa Indonesia · Tiếng Việt · ไทย · Français · Deutsch · Español · Português · Italiano · Русский · Türkçe

ux-skill — the design intelligence engine for Claude Code, Cursor, and every other AI coding tool

A design-intelligence engine that makes AI-generated UI distinctive instead of generic. Drop it into any of 17 AI coding tools and your output stops reading as AI-built. Free, MIT, offline, no LLM.

pip install uxskill

Star ux-skill on GitHub if this is useful — it is the single cheapest way to help the project. New here? Start with the 60-second tour or see it live at uxskill.laithjunaidy.com.

v3.1.0 — THE BRAIN, now brand-true + responsive. The strongest UX plugin for AI coding. A Python reasoning core with a deterministic 7-axis synthesizer, 12 queryable JSON manifests (84 styles, 176 palettes, 70 type pairings, 148 components, 184 industries, 35 chart types, 57 motion presets, 112 UX laws, 152 anti-pattern rules, 25 tech stacks, 160 brand specs), 25 slash commands, 5 sub-agents, 18 MCP tools, and a deterministic anti-AI-slop linter. Cross-IDE: ships into Claude Code, Cursor, Windsurf, GitHub Copilot, Gemini CLI, Codex, Kiro, Cline, Continue, Aider, Zed, JetBrains AI, Pieces, Tabby, Tabnine, CodeWhisperer, and Roo Cline.

The brand name is ux-skill. The PyPI / npm package name stays uxskill. The GitHub repo lives at Laith0003/ux-skill.

Site: uxskill.laithjunaidy.com · Compare vs every Claude UX plugin: compare.html · GitHub: Laith0003/ux-skill · PyPI: uxskill · npm: uxskill

New in v3.1 — brand-true, responsive, alive

Brand fidelity is enforced, not hoped. The primary color is read from the LOGO's pixels (not the most-painted CSS); default fonts are rejected for the logo's letterform style. The extracted brand travels recommend -> synthesize, and a hard floor in evaluate FAILS any output that drops the brand color/logo or ships no real imagery. Two-way interop with the open brand.md convention (render + ingest).
Mobile-first, gated. New craft foundations (responsive.md, component-behaviors.md) plus a wrap-aware gate that fails on horizontal scroll, a wrapping nav/wordmark/button label, or an over-tall sticky header.
The wow layer. The engine derives 2-3 coordinated signature moments per page — the "wow can only come from the user" doctrine is overturned.
Sharper linter (152 rules): imagery-mandatory + icon-only detection, placeholder-token and 100vw rules; seeded picsum kept, random stripped.

Tests 310 passing. Offline. Deterministic. No LLM ever called. Full notes in CHANGELOG.md.

What's new in v3

Brand specs become training data, not templates. The 160 brand specs are no longer a catalogue the recommender picks from — they're vocabulary the synthesizer distills from. Output is novel every call.
7-axis synthesizer (warmth, contrast, density, geometry, formality, motion, type_personality). Briefs map deterministically to axis values; axis values compile to fresh palette + type + spacing + radius + motion tokens.
Three auto-dispatched modes — strict_brand (100% of one brand), brand_anchor (70% one brand + 30% axis-adapted siblings), pure_synthesis (no brand named — distill from 8 axis-matching exemplars).
Decisions ledger drives the recommender. .ux/decisions.jsonl re-ranks candidates by past wins in the same (industry, ui_type) bucket. Cold-start safe. Counts only lint_score >= 80 + user_accepted = true decisions.
Axis interaction matrix — explicit conflict resolution between competing axes (dense + corporate → 4px, airy + corporate → 12px, soft + playful → 18px radius). No more silent ad-hoc rules.
/ux-evolve auto-loop — lint → polish → re-lint until score ≥ 90 or plateau or 5 rounds. Quality gate at 65.
3 new MCP tools (15 → 18): ux_synthesize, ux_decisions_query, ux_decisions_stats.
Local stats dashboard — uxskill stats --html writes .ux/stats.html showing what YOUR install has learned. No telemetry, no global aggregate.
223 tests pass. Offline. Deterministic. No LLM ever called.

Full details in CHANGELOG.md.

Star history

What is ux-skill

ux-skill is a design intelligence engine for AI coding tools. It runs as a Python package (pip install uxskill), as a Claude Code plugin, and as a 17-IDE multi-installer. The engine ingests a project brief (industry, audience, tone, must-haves, forbidden moves, stack, region) and returns a complete recommended design system: style, palette, type pair, motion presets, components, brand exemplars to study, and the anti-pattern guardrails that must hold. The recommendation is deterministic — same input always produces the same output.

The plugin sits between you and the AI coding tool. When you ask Claude Code, Cursor, or any other AI assistant to "build a fintech landing page," the assistant typically improvises — and the result reads as AI-generated within five seconds (purple-to-blue gradients, three equal cards, Inter at display size, "John Doe" in testimonials, 300ms default transitions, centered hero, bouncing arrow CTAs). ux-skill replaces improvisation with structured constraints: you run /ux-discover to capture the brief, /ux-recommend to pick the system, /ux-design to generate the code, and /ux-lint to verify it passes the 152 deterministic anti-AI-slop rules before commit.

This README is the canonical reference. Every command, every sub-agent, every data manifest, every install path, every brand spec, every anti-pattern category — it's all documented here. If you're shopping for a Claude Code design plugin or comparing AI design tools for Cursor, Windsurf, or Codex, read this top to bottom and the compare.html side by side.

The Brain — what v3.0 is
Quick install
The numbers — live comparison vs the top 8 Claude UX skills
Architecture — how the pieces fit
The 25 slash commands — detailed reference
The 5 sub-agents
The 11 data manifests
The 152 anti-AI-slop rules — the linter
The 160 brand DESIGN.md specs — by category
MCP server — the asymmetric move
The 17-IDE installer
Use cases — concrete scenarios
Compared to alternatives
Roadmap
Contributing
License, author, acknowledgments

The Brain — what v3.0 is

v3.0.0 is the biggest architectural shift in ux-skill's history. The recommender no longer picks templates from a catalogue — the engine synthesizes a fresh design language per brief. Same brief always yields the same output (fully deterministic), but every distinct brief gets its own novel system. Brand specs aren't templates anymore; they're training data the engine learns the vocabulary from. The system has eyes on its own history, closes the feedback loop locally, and never calls an LLM.

The compiler is a deterministic 7-axis synthesizer — warmth, contrast, density, geometry, formality, motion, type_personality. Every brief maps to axis values; axis values compile to fresh palette + type + spacing + radius + motion tokens. Modular type scales pick their ratio from contrast (1.200 quiet / 1.250 balanced / 1.333 loud). Layout primitives are responsive by construction (auto-fit minmax(min(N, 100%), 1fr) + container queries). Broken layouts can't be emitted because they aren't representable.

There are three auto-dispatched modes: strict_brand (reference_brands=[stripe] strict=True → 100% Stripe tokens, fastest path); brand_anchor (reference_brands=[stripe] → 70% Stripe + 30% axis-adapted from 4 sibling brands); and pure_synthesis (no brand named → infinity space, 8 axis-matching exemplars distilled into a novel design language). Conflicting axes are resolved by a documented axis interaction matrix — dense + corporate compiles to 4px (density wins, Bloomberg-school), airy + corporate to 12px (formality wins, luxury), soft + playful to 18px radius, sharp + corporate to 2px. No silent ad-hoc rules in the implementation.

The decisions ledger (.ux/decisions.jsonl, schema _v: 1 locked) closes the feedback loop. The recommender now re-ranks candidates by past wins in the same (industry, ui_type) bucket. Cold-start safe — it skips below 3 priors. It only counts decisions with lint_score >= 80 AND user_accepted = true. Plus /ux-evolve runs lint → polish → re-lint until score ≥ 90 or plateau or 5 rounds, with a 65-score quality gate below which output is refused unless --force. The result: every install gets smarter on its own corpus, every run is reproducible across machines, and the engine stays fully offline.

Quick install

Three install paths. Pick the one that matches your environment.

Path 1 — Claude Code marketplace (canonical)

If you live in Claude Code, install via the plugin marketplace:

/plugin marketplace add Laith0003/ux-skill
/plugin install ux@ux-skill

That wires all 25 slash commands and 5 sub-agents into your Claude Code session. After install, run /ux-init to set up the per-project .ux/ state directory and verify the Python engine is reachable.

Path 2 — pip (universal)

If you live outside Claude Code (Cursor, Windsurf, CLI, CI), install the Python package:

pip install uxskill
uxskill init                       # auto-detects your IDE, installs the right artifact
uxskill stats                      # print manifest counts to verify install
uxskill lint .                     # run the linter against the current directory

The package exposes both ux and uxskill as CLI entry points — they're the same binary.

Path 3 — npx (no Python required)

If you don't want to manage Python directly, the npx wrapper bootstraps everything via pipx:

npx uxskill init                  # downloads pipx + uxskill on first run
npx uxskill recommend --industry=fintech-neobank --tone=warm --stack=nextjs-15-app-router

Verify install

ux stats
# {
#   "version": "3.0.0-stable",
#   "counts": {
#     "styles": 84,
#     "palettes": 176,
#     "type-pairs": 70,
#     "components": 148,
#     "industries": 184,
#     "chart-types": 35,
#     "tech-stacks": 25,
#     "ux-guidelines": 112,
#     "motion-presets": 57,
#     "anti-patterns": 145,
#     "brands": 160
#   }
# }

If any count returns 0, the JSON file is missing — open an issue at github.com/Laith0003/ux-skill/issues.

The numbers — live comparison vs the top 8 Claude UX skills

Star counts last verified via gh api on 2026-05-28. ux-skill (Laith0003/ux-skill) is the newest entrant — we're tiny on awareness, deep on architecture. The comparison below is honest: where we lose, where we win.

Plugin	Stars	Architecture	Slash commands	Linter (CI-safe)	Brand specs	Components	Motion presets	IDEs supported
nextlevelbuilder/ui-ux-pro-max-skill	83,958	Python BM25 + CSV, single skill	1	—	—	0	0	18
nexu-io/open-design	54,406	Node.js + 19 skills + preview	19	—	—	0	0	1
Leonxlnx/taste-skill	25,202	Bash + research-backed taste	1	—	—	0	0	1
alchaincyf/huashu-design	15,455	Single 62 KB SKILL.md + scripts	1	—	—	0	0	1
google-labs-code/stitch-skills	5,762	MCP-wired skill library	multi	—	—	0	0	1
dominikmartn/nothing-design-skill	2,391	Single-aesthetic skill	1	—	—	0	0	1
Nutlope/hallmark	2,164	Anti-slop design skill	1	—	—	0	0	1
hamen/material-3-skill	955	MD3 components + audit	1	—	(MD3 only)	0	0	1
Laith0003/ux-skill (ux-skill)	14	Python engine + 12 manifests + 25 commands + 5 sub-agents + CI linter	22	152 regex rules	160	148	57	17

Where we lose

Awareness. They have hundreds of thousands of stars. We have 14. Star us — it's the cheapest way to help.
Brand recognition. ui-ux-pro-max and open-design have a head start measured in months, not days.
Marketing polish. They have screenshots, demo videos, and a discoverable landing page. We have a thorough README and a thin landing.

Where we win

Component library: 148 documented components with anatomy, states, tokens used, and motion specs. None of the other 8 ship a component manifest.
Motion presets: 57 stack-ready entries (Framer Motion, GSAP, CSS) with reduced-motion fallbacks. None of the others ship a motion manifest.
Anti-pattern linter: 152 deterministic regex rules, runs in CI, exits non-zero on Critical/High. None of the others ship a deterministic linter.
Brand specs: 160 real DESIGN.md specs (Apple, Stripe, Linear, Figma, Tesla, BMW, Notion, Spotify, Airbnb, Vercel, Supabase, Cursor, Raycast, Claude, and 96 more). None of the others ship a brand library.
17 IDEs supported: same engine, different glue per IDE.
25 slash commands: discovery, generation, audit, lint, polish, fix loop, case-study, workshop, copy, motion, a11y, dashboard, conductor — fully integrated.

Full table-by-table side-by-side at uxskill.laithjunaidy.com/compare.html.

Architecture — how the pieces fit

ux-skill (package name: uxskill)
│
├── data/                              The brain — queryable JSON manifests
│   ├── styles.json                    84 design styles + when/skip + tokens
│   ├── palettes.json                  176 palettes (light/dark, contrast verified)
│   ├── type-pairs.json                70 display × body × mono triplets
│   ├── components.json                148 components (anatomy, states, motion)
│   ├── industries.json                184 industry rules + audience signals
│   ├── chart-types.json               35 chart types (when/skip, encoding)
│   ├── tech-stacks.json               25 stacks (Next, Astro, SvelteKit, Blade...)
│   ├── ux-guidelines.json             112 named UX laws (Hick, Fitts, Miller...)
│   ├── motion-presets.json            57 motion presets (entry, exit, hover...)
│   ├── anti-patterns.json             152 regex rules (CI-safe linter source)
│   └── brands/*.json                  160 brand DESIGN specs + _index.json
│
├── engine/                            Python — the reasoning
│   ├── synthesizer/                   v3 — 7-axis deterministic compiler
│   ├── decisions/                     v3 — .ux/decisions.jsonl ledger + recommender re-rank
│   ├── recommender/                   5-parallel-search merge engine (re-ranked by decisions)
│   ├── linter/                        Deterministic anti-slop scanner
│   ├── discovery/                     10-field forcing protocol
│   ├── generator/                     Token + manifest emitter
│   ├── installer/                     17-IDE multi-installer
│   └── cli/                           `ux` / `uxskill` entry point
│
├── commands/                          22 Claude Code slash commands (.md)
│   ├── ux-init.md                     bootstrap
│   ├── ux-stats.md                    inventory snapshot
│   ├── ux-discover.md                 10-field intake (gate)
│   ├── ux-recommend.md                FLAGSHIP — 5-parallel search
│   ├── ux-lint.md                     deterministic linter
│   ├── ux-design.md                   generate frontend code
│   ├── ux-component.md                generate one component
│   ├── ux-system.md                   generate full design system
│   ├── ux-dashboard.md                generate dashboard surface
│   ├── ux-motion.md                   motion treatment + audit
│   ├── ux-audit.md                    6-lens design audit
│   ├── ux-a11y.md                     WCAG 2.1 AA audit
│   ├── ux-critique.md                 taste critique (3 wins, 3 misses, 1 move)
│   ├── ux-copy.md                     microcopy review + rewrite
│   ├── ux-fix.md                      apply findings as atomic commits
│   ├── ux-polish.md                   cosmetic pass + AI-slop kill
│   ├── ux-frame.md                    4-field framing block
│   ├── ux-research.md                 research planning + synthesis
│   ├── ux-workshop.md                 5-phase design thinking workshop
│   ├── ux-case-study.md               publishable Wfrah-editorial case study
│   ├── ux-next.md                     workflow conductor (read-only)
│   └── ux-expert.md                   consulting hook
│
├── agents/                            5 sub-agents (.md)
│   ├── frontend-engineer.md           React/Next/Vue/Blade/Astro
│   ├── motion-engineer.md             Framer Motion / GSAP / CSS
│   ├── copy-writer.md                 microcopy in brand voice
│   ├── research-synthesizer.md        interviews + analytics + competitors
│   └── design-system-architect.md     tokens / components / foundations
│
├── references/                        Prose source for the data + demo pages
│   ├── foundations/                   anti-patterns.md, principles, taste
│   ├── laws/                          UX laws long-form
│   ├── process/                       discovery-protocol.md (load-bearing)
│   ├── styles/                        per-style prose (anti-slop.md, etc.)
│   ├── components/                    component long-form
│   ├── output/                        output rubrics
│   └── conditional/                   stack-specific guidance
│
├── bin/
│   ├── uxskill.mjs                    npx wrapper -> Python engine
│   ├── ux-lint.py                     v2 linter (preferred)
│   └── ux-lint.sh                     v1 fallback (bash + perl-PCRE)
│
└── .ux/                               (created per project)
    ├── last-discovery.json            brief snapshot
    ├── last-recommendation.json       picked system
    ├── last-frame.json                framing block
    ├── last-audit.json / last-a11y.json / last-copy.json / last-motion.json
    ├── last-design.json / last-component.json / last-dashboard.json
    └── last-critique.json / last-polish.json / last-research.json / last-workshop.json / last-case-study.json

How the engine actually works

Input. You provide a brief — either interactively via /ux-discover (10 fields) or non-interactively via flags to ux recommend.
5 parallel searches. The engine runs five lookups concurrently across the manifests:
- Industry → recommended_styles (industries.json)
- Style → palette + type + motion compatibility (styles.json)
- Tone × must-have → palette filter (palettes.json)
- Stack → component compatibility + motion presets (tech-stacks.json, motion-presets.json)
- Forbidden + region → guardrails + brand exemplar shortlist (anti-patterns.json, brands/)
Merge. A deterministic merger ranks candidates, resolves conflicts (e.g., must-have dark-mode forces palette mode), and emits a single recommended system.
Output. A JSON document with the picked style, palette, type pair, top 5 motion presets, top 12 components, top 5 brand exemplars, and all 152 anti-pattern guardrails active. Plus a rationale block explaining each pick.
Generation. Downstream commands (/ux-design, /ux-component, /ux-system, /ux-dashboard) consume the recommendation to generate actual code via the sub-agents.
Verification. /ux-lint re-scans the generated code against the 152 regex rules. Exits non-zero on Critical/High in CI.

v3 additions. The recommender now re-ranks candidates from engine/decisions/ using .ux/decisions.jsonl (only counts decisions with lint_score >= 80 AND user_accepted = true; cold-start safe below 3 priors). The generator path can dispatch into engine/synthesizer/ — a deterministic 7-axis compiler that produces fresh palette + type + spacing + radius + motion tokens per brief instead of picking templates from a catalogue. See The Brain — what v3.0 is for details.

Python thinks. HTML shows. Markdown chains.

The 25 slash commands — detailed reference

Every command is shipped as a .md file under commands/ with description, allowed-tools, triggers, when to use, when to skip, input, process, and output state file. The descriptions below are condensed; the full source is the canonical spec.

Commands are grouped into five buckets: bootstrap & inventory, discovery & recommendation, generation, audit & verify, fix & polish, and conductor.

Bootstrap & inventory

`/ux-init` — bootstrap the project

What: Detects which IDE you're using (.claude/, .cursor/, .windsurf/, etc.), installs the right artifact, verifies the Python engine is reachable, prints a stats snapshot.
When to use: First time installing in a new project. After cloning a project that uses ux-skill. After pip install --upgrade uxskill.
When to skip: You already ran it in this project and nothing changed.
Invocation: /ux-init (no args) or uxskill init from the CLI.
Output: Per-IDE artifact (see The 17-IDE installer) + .ux/ directory + stdout summary.
Chains to: /ux-discover next.

`/ux-stats` — print the data inventory

What: Prints version + entry counts for the 11 data manifests, so you can verify what's installed.
When to use: After install. After upgrade. When /ux-recommend returns surprising picks and you suspect the manifests are incomplete.
When to skip: Never — it's a 50ms read-only command.
Invocation: /ux-stats or uxskill stats.
Output: JSON to stdout (see Verify install above).
Chains to: Diagnostic only; doesn't feed downstream.

Discovery & recommendation

`/ux-discover` — the forcing function (10-field intake)

What: The mandatory 10-field intake every project goes through before any generation command. Project type, audience, primary goal, tone, must-haves, forbidden, reference brands, stack, region, success metric. No improvisation. Banned phrases ("modern", "clean") force the user to be specific.
When to use: Before any /ux-design, /ux-component, /ux-system, or /ux-dashboard. Whenever a previous brief has gone stale.
When to skip: You're fixing a bug (/ux-fix). You're only running a linter pass (/ux-lint). The brief is unchanged from the last session.
Invocation: /ux-discover. The plugin asks; you answer.
Output: Writes .ux/last-discovery.json (the 10-field brief).
Chains to: /ux-recommend → uses the discovery to pick style + palette + type + motion + components. /ux-design [extra brief] → generates frontend code grounded in the recommendation. /ux-component <name> → generates one component aligned to discovered constraints.

`/ux-recommend` — the flagship 5-parallel-search engine

What: Runs the Python engine's 5-parallel-search across 12 manifests and returns one merged design system. Industry → Style → Palette → Type → Motion + Components + Brand Exemplars + Guardrails.
When to use: Starting a new project from zero. Pivoting a tired-looking product. Pre-flight before any /ux-design or /ux-component.
When to skip: You already ran /ux-discover and saved a brief — /ux-recommend is automatic in that flow. You're fixing one bug (use /ux-fix). You only need to lint (use /ux-lint).

Invocation (Claude Code):

/ux-recommend

Invocation (CLI):

ux recommend \
  --project-type=landing \
  --industry=fintech-neobank \
  --tone=warm --tone=editorial \
  --must-have=dark-mode --must-have=a11y-AA \
  --forbidden=brutalism --forbidden=purple-gradients \
  --stack=nextjs-15-app-router \
  --region=mena

Output: Writes .ux/last-recommendation.json — picked style, picked palette, picked type pair, top 5 motion presets, top 12 components, top 5 brand exemplars, all 152 anti-pattern guardrails active, plus rationale.
Chains to: /ux-design [brief] → frontend code using the recommended tokens. /ux-system → full design system from the recommendation. /ux-component <name> → one component using the recommended style. /ux-lint → verify the generated code.

Generation

`/ux-design` — generate a beautiful, anti-slop surface from a brief

What: Generates a complete, production-grade frontend artifact (landing, marketing site, app shell) from the discovery brief + recommendation. Dispatches frontend-engineer with creative direction from anti-slop and arsenal references.
When to use: "Design a", "build me a", "generate a landing page", "create a dashboard", "make a component" — any free-form visual deliverable request.
When to skip: You want a review, not a build (use /ux-audit or /ux-critique). You want one component only (use /ux-component). Backend or infrastructure work.
Invocation: /ux-design generate a fintech landing for a MENA neobank, warm editorial tone, dark-mode AA, no purple gradients.
Output: Generated code (HTML / Blade / JSX / Vue / Astro), plus .ux/last-design.json.
Chains to: /ux-lint → verify against guardrails. /ux-polish → cosmetic pass. /ux-a11y → accessibility audit. /ux-copy → microcopy review. /ux-fix → apply findings as atomic commits.

`/ux-component` — generate one component

What: Produces a single, production-grade component (button, modal, navbar, sidebar, card, table, form, chart) from a spec. All four interaction states, accessible, on-brand. Looks up the component in .ux/last-recommendation.json first, falls back to direct manifest query.
When to use: Any single-element request — "build a button", "create a pricing card", "make a modal", "add a navbar", "design a sidebar", "I need a data table", "build a form", "make a chart component".
When to skip: Full page or multi-section surface (use /ux-design). Backend or infrastructure.
Invocation: /ux-component pricing-card-trio --brief="fintech, dark, monospace numbers".
Output: Generated component code, plus .ux/last-component.json.
Chains to: /ux-lint → verify. /ux-polish → tighten.

`/ux-system` — generate a complete starter design system

What: Proposes a complete starter design system for a project that doesn't have one — tokens (color, type, space, motion, radius, shadow), foundation docs, component contracts, dark-mode pairings, theme switcher. Dispatches design-system-architect.
When to use: "We don't have a design system", "build us a system", "propose tokens", "what should our theme be", "set up our DS".
When to skip: The project already has a design system — use /ux-component against the existing system instead. Backend or infrastructure.
Invocation: /ux-system (runs discovery first if not already on file).
Output: tokens.json, foundations.md, components/*.md contracts, optional Tailwind / vanilla / SCSS emit. Writes .ux/last-system.json for chain context.
Chains to: /ux-component → build against the new system. /ux-design → generate a surface using the new tokens.

`/ux-dashboard` — specialized dashboard generation

What: Dashboard with data density discipline — bento layout, tabular monospace numerals, sparkline patterns, anti-card-overuse, semantic state colors, sparing motion. Not a marketing site with charts pasted on.
When to use: "Build a dashboard", "design the admin panel", "make a metrics page", "operator console", "analytics view", "KPI board", "monitoring screen".
When to skip: Marketing landing page with stats (use /ux-design). One widget only (use /ux-component). Backend or infrastructure.
Invocation: /ux-dashboard.
Output: Generated dashboard code + .ux/last-dashboard.json.
Chains to: /ux-lint, /ux-audit, /ux-a11y.

`/ux-motion` — motion treatment

What: Generates the motion layer of a surface — durations, easings, choreography, reduced-motion fallbacks, performance discipline. Also audits existing motion against the 5 dimensions (timing, easing, meaning, reduced-motion, performance).
When to use: "Motion check", "are the animations good", "fix the motion", "review the animations", "motion audit", "performance pass on the motion".
When to skip: Surface has no motion (use /ux-audit or /ux-polish). Backend or infrastructure.
Invocation: /ux-motion path/to/component.tsx (audit mode) or /ux-motion --generate hero-entry (generation).
Output: Updated code (in generation mode) or .ux/last-motion.json report (in audit mode).
Chains to: /ux-fix → apply motion findings. /ux-polish → tighten.

Audit & verify

`/ux-lint` — deterministic regex-based linter (no LLM, CI-safe)

What: Runs 152 regex rules against your code. No LLM call. Exits non-zero on Critical / High in CI. Source: data/anti-patterns.json. Rules cover A11y (23), Content (15), Layout (13), Typography (10), Color (9), Quality (9), Visual (9), Motion (8), Performance (4).
When to use: Pre-commit hook. CI gate. Fast first pass on a large codebase before paying the cost of /ux-audit. After /ux-design or /ux-component to verify generation.
When to skip: You want a fix loop (the linter reports, it does not edit — chain into /ux-polish --fix or /ux-fix). You want taste judgment (use /ux-critique).
Invocation (slash): /ux-lint src/.
Invocation (CLI): uxskill lint . or python3 bin/ux-lint.py . or bash bin/ux-lint.sh --ci --fail-on high.

Invocation (CI):

- name: ux-lint
  run: bash bin/ux-lint.sh --ci --fail-on high

Output: Findings to stdout (location, rule id, severity, evidence). Exit code 0 if clean, non-zero on Critical/High when --fail-on high is set.
Chains to: /ux-polish --fix → LLM-driven counterpart on the same patterns. /ux-fix → apply findings as commits, severity-sorted. /ux-audit → full 6-lens reasoning pass. /ux-next → let the conductor decide.

`/ux-audit` — 6-lens design audit

What: A structured, opinionated review against six lenses (clarity, hierarchy, accessibility, voice, motion, taste), producing severity-tagged findings. Polaris-style report. Reads .ux/last-frame.json first — audience and outcome anchor every finding's severity.
When to use: Surface exists and you want a defensible critique. "Audit", "review the ux", "is this any good", "what's broken", "tear this apart".
When to skip: Surface doesn't exist yet (use /ux-design). User wants one lens (use the targeted command: /ux-a11y, /ux-copy, /ux-motion, /ux-polish). User wants taste opinion (use /ux-critique). Backend or infrastructure.
Invocation: /ux-audit https://example.com/pricing or /ux-audit src/components/Pricing.tsx.
Output: Writes .ux/last-audit.json — findings array of {lens, severity, title, principle, evidence, fix}, severity_counts, dominant_lens, strategic_moves.
Chains to: /ux-fix → apply findings. /ux-polish → cosmetic pass. /ux-design → if structural redesign needed.

`/ux-a11y` — WCAG 2.1 AA audit + common-courtesy checks

What: A structured WCAG 2.1 AA audit, plus the common-courtesy checks that pass automated tools but still hurt real users (focus visibility, error specificity, motion preferences, keyboard traps, color reliance).
When to use: Pre-ship accessibility gate. After a redesign. "Accessibility check", "WCAG audit", "is this accessible", "a11y review", "screen reader test", "keyboard nav check".
When to skip: Not user-facing. Backend or infrastructure. Work-in-progress sketches.
Invocation: /ux-a11y https://example.com (live URL preferred — automated tools and keyboard testing only work live).
Output: Writes .ux/last-a11y.json — findings array of {wcag_sc, sc_name, severity, title, evidence, fix, category}, beyond_wcag array, severity_counts.
Chains to: /ux-fix → apply findings as commits. /ux-copy → fix alt text and form-error wiring as part of a copy pass.

`/ux-critique` — taste call (3 wins, 3 misses, 1 strategic move)

What: A designer's opinion — not a structured audit, not a severity score, just a tight, opinionated take that names what's working, what's not, and the one strategic move that would change the most.
When to use: "What do you think", "is this good", "critique this", "honest take", "is the vibe right", "does this feel like us", "should we ship this".
When to skip: User explicitly wants a structured audit (use /ux-audit). Backend or infrastructure.
Invocation: /ux-critique https://example.com.
Output: Writes .ux/last-critique.json — 3 wins, 3 misses, 1 strategic move, plus prose.
Chains to: /ux-design if the take recommends redesign. /ux-polish if the take recommends tightening.

`/ux-copy` — microcopy review + rewrite

What: Evaluates every visible string against the voice rubric and produces a before/after rewrite. Catches: "form contains errors" (generic), "John Doe" (placeholder), AI-cheerful celebratory copy, generic CTAs, dead empty states, useless errors.
When to use: Structure is right but words are weak. "Review the copy", "fix the microcopy", "the error messages are bad", "rewrite this", "tighten the strings", "the buttons sound generic", "this empty state is dead".
When to skip: Layout problems (use /ux-audit or /ux-polish). Accessibility-driven copy issues like alt text (use /ux-a11y). Backend or infrastructure.
Invocation: /ux-copy src/views/checkout.blade.php.
Output: Writes .ux/last-copy.json — strings array of {location, severity, before, after, notes}, plus rubric + locales needing translation.
Chains to: /ux-fix → apply rewrites. /ux-a11y → re-check after copy fixes.

Fix & polish

`/ux-fix` — apply findings as atomic commits

What: Reads the latest report from .ux/ (audit, copy, a11y, motion, or polish), validates the working tree, and applies findings as atomic commits via the right sub-agents. Re-verifies by re-running the originating command.
When to use: After running an audit-class command and reviewing findings. "Fix the findings", "apply the fixes", "run the fix loop", "patch the surface", "make the changes", "go fix it".
When to skip: No prior report in .ux/. Working tree is dirty and the user hasn't agreed to stash/commit. Fixes need design judgment, not mechanical application (use /ux-design for a redesign).
Invocation: /ux-fix (auto-detects which report to fix) or /ux-fix --from=last-a11y.json.
Output: Atomic commits per finding. Re-runs the originating command and updates the .ux/last-*.json file. Prints a summary.
Chains to: /ux-next → conductor picks the next move.

`/ux-polish` — cosmetic pass + AI-slop kill

What: Spacing rhythm, hierarchy sharpening, AI-slop detection, token consistency. The LLM-driven counterpart to /ux-lint — uses your judgment on taste calls.
When to use: Structure is right but execution is loose. "Polish", "tighten this up", "remove the AI-slop", "make it premium", "make this less AI-looking", "the spacing feels off", "this looks generic", "needs more taste".
When to skip: Surface is missing core functionality (fix that first). Needs a redesign, not a polish (use /ux-design). Copy issues (use /ux-copy). Motion issues (use /ux-motion). A11y issues (use /ux-a11y).
Invocation: /ux-polish src/components/Hero.tsx.
Output: Updated code + .ux/last-polish.json describing the changes.
Chains to: /ux-lint → verify the polish held. /ux-a11y → re-check accessibility.

Discovery & narrative

`/ux-frame` — 4-field framing block

What: Captures who-it's-for, outcome, hypothesis, and success signal in a structured framing block. No design work happens — just the four-field intake that turns a vague request into a working brief. Lighter than /ux-discover (4 fields vs 10).
When to use: Start of any project, sprint, or one-off engagement. Mid-stream when a conversation has drifted. "Frame this", "what's the brief", "set up the project", "framing".
When to skip: Already framed (check .ux/last-frame.json). One-off component build with no framing implications. Backend or infrastructure.
Invocation: /ux-frame "loyalty wallet for MENA Bashiti pilot".
Output: Writes .ux/last-frame.json — {audience, outcome, hypothesis, success_signal}.
Chains to: /ux-discover → extend the frame to the 10-field brief. /ux-design → generate using the frame as anchor.

`/ux-research` — research planning + synthesis

What: Planning mode: writes interview scripts, surveys, recruitment screeners. Synthesis mode (--synthesize): digests interviews, analytics, competitor sites, A/B results, support tickets into recommendations. Dispatches research-synthesizer.
When to use: "Plan a research study", "I need interview questions", "design a survey", "how do I recruit users", "user testing plan", "diary study", "preference test", "fake door", "smoke test", "synthesize my interview notes".
When to skip: Answer is already known with high confidence. Low-risk reversible decisions. Backend or infrastructure.
Invocation: /ux-research --plan "loyalty wallet adoption in MENA" or /ux-research --synthesize interviews/*.md.
Output: Writes .ux/last-research.json — research plan or synthesized themes + evidence + recommendations.
Chains to: /ux-frame → integrate findings into a frame. /ux-design → generate from findings. /ux-workshop → run a workshop using the research as input.

`/ux-workshop` — 5-phase design thinking workshop

What: Facilitates a discovery / design-thinking workshop end-to-end. Five sequential phases (exploration → heat map → stakeholder map → solution sketch → game plan). Time-boxed. Concrete artifacts per phase. Ends with a decision, not "interesting findings."
When to use: Real question, real participants, real time budget. "Run a workshop", "facilitate a discovery", "let's do a design thinking session", "I have stakeholders for an hour, what do we do", "kick off the project".
When to skip: Brief is already clear and scoped. Solo brainstorm (use /ux-design or /ux-frame). Team is mid-execution, not in discovery.
Invocation: /ux-workshop "loyalty wallet pivot" --participants="2 PMs, 1 designer, 1 eng lead, 1 customer rep" --minutes=90.
Output: Writes .ux/last-workshop.json — game plan + per-phase artifacts.
Chains to: /ux-design → execute the game plan. /ux-research → fill gaps the workshop surfaced. /ux-case-study → publish the journey.

`/ux-case-study` — publishable case study (Wfrah-editorial format)

What: Generates a project case study in pure-monochrome editorial format — Wfrah typography, hairline separators, numbered (A)–(G) section codes, bilingual-safe layout. A document, not a marketing brochure. Reads from .ux/last-frame.json, .ux/last-workshop.json, .ux/last-research.json, .ux/last-design.json, .ux/last-a11y.json, .ux/last-polish.json, .ux/last-recommendation.json, .ux/last-discovery.json.
When to use: Post-launch. After a discrete milestone. "Write a case study", "case study this project", "do the wrap-up doc", "publish this work", "portfolio piece".
When to skip: Project lacks data to populate (A)–(G) sections. User wants a marketing landing, not a case study (use /ux-design).
Invocation: /ux-case-study --format=html --slug=bashiti-loyalty.
Output: case-studies/<slug>.<ext> + .ux/last-case-study.json.
Chains to: Terminal command — usually the end of a project.

Conductor

`/ux-next` — workflow conductor (read-only)

What: Reads every .ux/last-*.json and names the highest-leverage next command. A conductor, not a builder. Read-only.
When to use: Between commands. "What should I do next", "what's the next move", "decide for me", "where do we go from here".
When to skip: No prior reports in .ux/. You have a specific next command in mind.
Invocation: /ux-next (no args) or /ux-next --focus=a11y.
Output: Stdout — recommended next command + rationale.
Chains to: Whichever command it picks.

`/ux-expert` — consulting hook

What: Surfaces the plugin creator's contact info when a user asks for a real-life UX expert. Brief, direct, no marketing.
When to use: "Who built this", "I need a UX expert", "do you do consulting", "can I hire someone for this", "is there a human behind this plugin".
When to skip: User is asking about plugin features, not about consulting.
Invocation: /ux-expert.
Output: Brief contact card with LinkedIn / email / repo.

Command chaining graph

                  ┌──────────────────────┐
                  │  /ux-init            │
                  │  /ux-stats           │
                  └────────────┬─────────┘
                               │
                  ┌────────────▼─────────┐
                  │  /ux-frame           │  4-field framing block
                  └────────────┬─────────┘
                               │
                  ┌────────────▼─────────┐
                  │  /ux-discover        │  10-field intake (FORCING GATE)
                  └────────────┬─────────┘
                               │ writes .ux/last-discovery.json
                  ┌────────────▼─────────┐
                  │  /ux-recommend       │  5 parallel searches -> merged system
                  └────────────┬─────────┘
                               │ writes .ux/last-recommendation.json
            ┌──────────────────┼──────────────────┐
            │                  │                  │
   ┌────────▼───────┐ ┌────────▼────────┐ ┌──────▼──────┐
   │ /ux-design     │ │ /ux-component   │ │ /ux-system  │
   │ /ux-dashboard  │ │ /ux-motion      │ │             │
   └────────┬───────┘ └────────┬────────┘ └──────┬──────┘
            │                  │                  │
            └──────────────────┼──────────────────┘
                               │ writes .ux/last-<surface>.json
            ┌──────────────────┼──────────────────┐
            │                  │                  │
   ┌────────▼───────┐ ┌────────▼────────┐ ┌──────▼──────┐
   │ /ux-lint       │ │ /ux-audit       │ │ /ux-a11y    │
   │ /ux-critique   │ │ /ux-copy        │ │ /ux-motion  │
   └────────┬───────┘ └────────┬────────┘ └──────┬──────┘
            │                  │                  │
            └──────────────────┼──────────────────┘
                               │ writes .ux/last-<lens>.json
                  ┌────────────▼─────────┐
                  │  /ux-fix             │  apply findings as commits
                  │  /ux-polish          │
                  └────────────┬─────────┘
                               │
                  ┌────────────▼─────────┐
                  │  /ux-case-study      │  publishable artifact
                  └──────────────────────┘

                  ┌──────────────────────┐
                  │  /ux-next            │  conductor — read-only
                  │  /ux-expert          │  consulting hook
                  └──────────────────────┘

The 5 sub-agents

Sub-agents are role-specific generators dispatched by commands. They never run independently — they're called by /ux-design, /ux-component, /ux-system, /ux-fix, /ux-research, etc. Each agent has a defined ownership boundary: they do NOT decide the brief; they execute against it.

`frontend-engineer`

Owns: Production-grade frontend code (React, Next.js, Vue, Blade+Alpine, vanilla HTML, Astro) with anti-AI-slop discipline.
Dispatched by: /ux-design, /ux-component, /ux-dashboard, /ux-fix.
Inputs: Brief + creative direction + tokens (from .ux/last-recommendation.json).
Outputs: Working code that's distinguishable from generic AI output. No purple gradients, no centered hero, no three equal cards, no Inter at display size, no "John Doe", no emoji, no 300ms defaults.
Tools: Read, Write, Edit, Bash, Glob, Grep.

`motion-engineer`

Owns: Motion in production frontend code — Framer Motion, GSAP, CSS animations. Durations, easings, choreography, reduced-motion fallbacks, performance discipline.
Dispatched by: /ux-design, /ux-motion --fix, /ux-component.
Inputs: Motion brief + tokens + the 57 motion presets from data/motion-presets.json.
Outputs: Motion that earns its place. Always wrapped in prefers-reduced-motion fallbacks. Always tested against Core Web Vitals.
Tools: Read, Write, Edit, Bash, Glob, Grep.

`copy-writer`

Owns: The strings that ship — error messages, empty states, CTAs, loading states, success messages, toasts, helper text, form labels, button text.
Dispatched by: /ux-copy --fix, /ux-design, /ux-frame, /ux-component.
Inputs: Voice profile (named or pasted) + the surface's strings.
Outputs: Production microcopy applied consistently across every state of a surface so the product sounds like one product, not ten. Bans: "form contains errors", "John Doe", AI-cheerful celebratory copy, generic CTAs, dead empty states.
Tools: Read, Write, Edit, Bash, Glob, Grep.

`research-synthesizer`

Owns: Digesting research inputs (interviews, analytics, competitive sites, A/B results, support tickets) into actionable design recommendations.
Dispatched by: /ux-research, /ux-workshop, /ux-frame.
Inputs: Raw research — transcripts, exports, competitor URLs, support clusters.
Outputs: Themes, evidence, recommendations. Never designs the answer — gives the designer the substrate to design from.
Tools: Read, Write, WebFetch, Bash, Glob, Grep.

`design-system-architect`

Owns: Complete design systems — tokens (color, type, space, motion, radius, shadow), foundation docs, component contracts, dark-mode pairings, theming layer.
Dispatched by: /ux-system, /ux-component when no system exists.
Inputs: Brand brief + .ux/last-recommendation.json (style + palette + type pair + motion presets).
Outputs: A coherent, opinionated, production-ready system that downstream agents can build against without re-deciding fundamentals. Tokens JSON, foundations MD, component contracts, dark-mode mapping.
Tools: Read, Write, Edit, Bash, Glob, Grep.

Sub-agent dispatch protocol

When a command dispatches a sub-agent, it passes:

The brief / recommendation (loaded from .ux/).
The relevant manifest slice (e.g., frontend-engineer gets the picked style + palette + components; motion-engineer gets the picked motion presets).
The 152 anti-pattern guardrails (always active).
A success criterion (what the artifact must do).

Sub-agents return:

The artifact (code, doc, system).
A rationale block (why these picks).
A self-check against the guardrails (which rules they verified).

The calling command then runs /ux-lint automatically before declaring done.

The 11 data manifests

The data layer is the brain. Every command reads from it; the engine merges across it; the linter scans against it. All files live under data/ and wrap their entries in {_meta, entries} for schema versioning.

`styles.json` — 84 design styles

Field	Description
`entries`	84
`keys per entry`	`id`, `name`, `category`, `philosophy`, `when_to_use`, `when_to_skip`, `tokens`, `references`, `compatible_palettes`, `compatible_type_pairs`, `compatible_motion`, `compatible_industries`, `taste_score`
`categories`	Minimalist / Swiss, Brutalist, Editorial, Glassmorphism, Neumorphism, Bento, Skeuomorphic, Industrial, Maximalist, AI-Futurist, MENA-modern, Vaporwave, etc.
`sample entry`	`swiss-international` — "Grid is law. Type does the heavy lifting. Decoration is failure."

Used by: /ux-recommend, /ux-system, /ux-design. Schema: data/SCHEMAS.md.

`palettes.json` — 176 color palettes

Field	Description
`entries`	176
`keys per entry`	`id`, `name`, `mode` (light/dark), `tone`, `colors` (canvas, surface, ink, body, muted, primary, primary_active, hairline, success, warning, danger, accent), `wcag_contrast_audit`, `compatible_industries`
`tones`	warm, editorial, magazine, clinical, playful, brutalist, monochrome, jewel-tone, MENA-warm, dev-tools-dark, etc.
`sample entry`	`claude-warm-editorial` — light, warm/editorial/magazine, canvas #faf9f5, primary #cc785c

Used by: /ux-recommend, /ux-system. Contrast verified at AA / AAA. Schema: data/SCHEMAS.md.

`type-pairs.json` — 70 type pairings

Field	Description
`entries`	70
`keys per entry`	`id`, `name`, `display` (family + weights + source + license + URL), `body`, `mono`, `compatible_styles`, `taste_score`
`sample entry`	`cormorant-inter-jetbrains` — Cormorant Garamond × Inter × JetBrains Mono

All families have license + source URL. Used by /ux-recommend, /ux-system.

`components.json` — 148 components

Field	Description
`entries`	148
`keys per entry`	`id`, `name`, `category`, `purpose`, `anatomy`, `states`, `tokens_used`, `motion`, `accessibility`, `compatible_styles`, `compatible_industries`, `code_skeleton`
`categories`	Navigation, Forms, Data Display, Feedback, Overlays, Layout, Content, Marketing, E-commerce, Auth, Dashboard, Charts, Empty States, Loading States, Error States
`sample entry`	`mega-nav-product-grid` — Mega Navigation, Product Grid — 6-part anatomy, 4 states

This is our biggest moat. No other Claude UX plugin ships a structured component manifest.

`industries.json` — 184 industry rules

Field	Description
`entries`	184
`keys per entry`	`id`, `name`, `category`, `characteristics`, `audience_signals`, `recommended_styles`, `recommended_palettes`, `recommended_type_pairs`, `recommended_motion`, `regulatory_notes`, `regional_notes`
`categories`	Financial Services, Healthcare, Education, E-commerce, SaaS B2B, SaaS B2C, Developer Tools, Media, Gaming, Travel, Real Estate, MENA-specific, etc.
`sample entry`	`fintech-neobank` — high trust, regulatory disclosures, balance/transaction primary UI, mobile-first daily-use

Used by /ux-recommend as the first parallel search axis.

`chart-types.json` — 35 chart types

Field	Description
`entries`	35
`keys per entry`	`id`, `name`, `category`, `when_to_use`, `when_to_skip`, `encoding`, `accessibility`, `data_shape`, `compatible_styles`
`categories`	Comparison, Time Series, Distribution, Composition, Relationship, Flow, Geographic
`sample entry`	`bar-vertical` — Compare 4–15 discrete categories. Position along x-axis maps category; height maps value.

Used by /ux-dashboard, /ux-component (chart instances).

`tech-stacks.json` — 25 stacks

Field	Description
`entries`	25
`keys per entry`	`id`, `name`, `category`, `tier`, `languages`, `ssr`, `rsc`, `compatible_styling`, `scaffold_command`, `compatible_motion`, `gotchas`
`tiers`	production, prerelease, experimental
`sample entry`	`nextjs-15-app-router` — Next.js 15 (App Router), TS/JS, SSR, RSC, compatible with Tailwind 4 / CSS Modules / vanilla-extract / styled-components / panda-css

Other stacks include Astro, SvelteKit, Remix, Nuxt 3, Solid Start, Qwik, Blade+Alpine, Hotwire, Phoenix LiveView, Hydrogen 2025.

`ux-guidelines.json` — 112 named UX laws

Field	Description
`entries`	112
`keys per entry`	`id`, `name`, `category`, `source`, `principle`, `application`, `examples`, `caveats`, `related_laws`
`categories`	Decision Cost, Attention, Memory, Motor Control, Visual Perception, Social, Emotional, Form, Error Handling, Onboarding, Empty State, etc.
`sample entry`	`hicks-law` — Decision time grows logarithmically with the number of choices presented

Used by /ux-audit (6-lens scoring) and /ux-critique (taste anchor).

`motion-presets.json` — 57 motion presets

Field	Description
`entries`	57
`keys per entry`	`id`, `name`, `category`, `tokens` (duration_ms, easing, transform_from/to, opacity_from/to), `stacks` (framer_motion, gsap, css), `accessibility` (reduced-motion fallback), `when_to_use`
`categories`	Entry, Exit, Hover, Focus, Tap, Loading, Empty, Success, Error, Scroll-linked
`sample entry`	`fade-up-12px` — 360ms, `cubic-bezier(0.16, 1, 0.3, 1)`, translateY(12px) → 0, opacity 0 → 1

Every preset has a reduced-motion variant. Stack-ready code for Framer Motion, GSAP, and pure CSS.

`anti-patterns.json` — 152 regex rules

Field	Description
`entries`	145
`keys per entry`	`id`, `name`, `severity` (critical/high/medium/low), `category`, `detection` (type, pattern, flags, scope), `evidence_template`, `fix`, `references`
`categories`	A11y (23), Content (15), Layout (13), Typography (10), Color (9), Quality (9), Visual (9), Motion (8), Performance (4)

The full rule list is in The 152 anti-AI-slop rules.

`brands/*.json` — 160 brand specs

Field	Description
`entries`	160 (plus `_index.json` listing all)
`keys per entry`	`id`, `name`, `category`, `voice`, `tokens` (color, type, motion), `design_principles`, `signature_moves`, `anti-moves`, `references`
`categories`	Developer Tools (36), Consumer / Lifestyle / Retail (19), Fintech / Crypto (14), Editorial / Media (13), AI / ML Platform (12), Productivity / Collaboration (8), Automotive (8)

Full list in The 160 brand DESIGN.md specs.

The 152 anti-AI-slop rules — the linter

ux-skill ships a deterministic regex-based linter. No LLM. No API. No network. Runs in CI in ~200ms on a typical Next.js app. Exits non-zero on Critical / High findings when --fail-on high is set.

The rules are sourced from data/anti-patterns.json (v2 preferred) with a references/foundations/anti-patterns.md fallback (v1 bash). Two binaries ship: bin/ux-lint.py (Python, fast, extensible) and bin/ux-lint.sh (Bash + perl-PCRE, for environments without Python).

Rules by category

Typography (3 rules)

Severity	Rule ID	Name
high	`inter-as-display`	Inter used as display font
medium	`hero-text-arbitrary-90px`	Arbitrary hero font size
low	`font-system-only`	System font stack with no chosen typeface

Color (6 rules)

Severity	Rule ID	Name
high	`purple-to-blue-gradient`	Default purple-to-blue AI gradient
high	`dark-text-on-dark-card`	Low-contrast text on card
medium	`gradient-text-rainbow`	Multi-stop gradient text
medium	`card-glow-purple-shadow`	Purple glow shadow on cards
medium	`gradient-mesh-purple-pink`	Purple-pink mesh gradient hero
low	`tailwind-color-named-vague`	Named Tailwind colors with no semantic token

Layout (5 rules)

Severity	Rule ID	Name
high	`three-equal-card-grid`	Three equal cards in a row
medium	`centered-everything-hero`	Centered hero composition
medium	`avatar-stack-overlapping`	Generic overlapping avatar stack
low	`pill-rounded-full-everywhere`	`rounded-full` applied to everything
low	`nav-equal-hamburger-desktop`	Hamburger menu on desktop

Content (5 rules)

Severity	Rule ID	Name
high	`lorem-ipsum-leak`	Lorem ipsum in shipping code
high	`emoji-in-ui`	Emoji used as UI element
high	`icon-emoji-stamp`	Emoji used as icon stamp
high	`testimonial-fake-five-stars`	Hardcoded five-star testimonial
medium	`fake-name-john-doe`	Generic placeholder names

Motion (3 rules)

Severity	Rule ID	Name
medium	`cta-arrow-rightward-bouncing`	Bouncing arrow on CTA
low	`timing-300ms-default`	Default 300ms transition timing
low	`cubic-bezier-material-only`	Material default easing everywhere

Severity	Rule ID	Name
high	`inline-svg-no-aria`	SVG without aria-label or aria-hidden
high	`img-no-alt`	Image missing alt attribute
high	`link-onclick-no-href`	Anchor with onClick but no href
medium	`button-no-type`	Button missing type attribute
medium	`heading-skip-h1-h3`	Skipped heading level
medium	`infinite-scroll-no-pagination`	Infinite scroll without keyboard fallback

Quality (6 rules)

Severity	Rule ID	Name
high	`console-log-leak`	`console.log` in component code
medium	`inline-style-attribute`	Inline style attribute
medium	`any-type-leak`	TypeScript `any` type
medium	`arbitrary-z-index-9999`	Lazy z-index value
low	`shadcn-default-everywhere`	Default shadcn token block unmodified
low	`todo-fixme-comment`	TODO or FIXME in shipping code

Visual (1 rule)

Severity	Rule ID	Name
low	`blur-bg-only-decoration`	Backdrop blur with no glass surface

Linter usage

One-off scan:

uxskill lint .
# or
python3 bin/ux-lint.py src/
# or
bash bin/ux-lint.sh src/

CI gate (GitHub Actions):

- name: ux-lint
  run: bash bin/ux-lint.sh --ci --fail-on high

Pre-commit hook:

# .git/hooks/pre-commit
#!/usr/bin/env bash
bash bin/ux-lint.sh --staged --fail-on high

Output (sample):

─── /ux-lint report ───
src/components/Hero.tsx:24  [high]   purple-to-blue-gradient
  evidence: bg-gradient-to-br from-purple-500 to-blue-500
  fix: replace with the recommended palette's primary gradient or remove gradient

src/components/Pricing.tsx:11  [high] three-equal-card-grid
  evidence: grid grid-cols-3 gap-6 (3 equal Card children)
  fix: feature one card; flank with two reduced-emphasis cards

3 files scanned · 2 high · 0 medium · 0 low · exit 1
Recommended next: /ux-polish --fix (LLM-driven, addresses both lintable and aesthetic findings)

The 160 brand DESIGN.md specs — by category

Real brands. Real design languages. Real DESIGN.md specs — not generic palettes. Tell the plugin "build a landing in Stripe's style" and it reads the actual brand vocabulary: voice rubric, color tokens, motion conventions, signature moves, anti-moves.

Each brand ships as a structured JSON (data/brands/<slug>.json) plus a prose reference (references/brands/<slug>.md).

Developer Tools (36)

ClickHouse, Composio, Cursor, Datadog, dbt Labs, Expo, Fivetran, Fly.io, Framer, HashiCorp, Honeycomb, IBM, Lovable, Mintlify, Modal, MongoDB, Neon, Ollama, OpenCode, PostHog, Railway, Raycast, Render, Replicate, Resend, Retool, Sanity, Sentry, Slack, Snowflake, Sourcegraph, Supabase, Superhuman, Vercel, Warp, Webflow

Consumer / Lifestyle / Retail (19)

Aesop, Airbnb, Allbirds, Apple, Apple Music, Glossier, HP, Hims & Hers, Instagram, Meta, Nike, Patagonia, Pinterest, PlayStation, Shopify, Spotify, Starbucks, TikTok, Uber

Fintech / Crypto (14)

Binance, Brex, Coinbase, Kraken, Mastercard, Mercury, Monzo, N26, Plaid, Ramp, Revolut, Robinhood, Stripe, Wise

Editorial / Media (13)

Bloomberg, Clay, Dezeen, NVIDIA, Pitchfork, Substack, The Atlantic, The Economist, The New York Times, The Verge, The Wall Street Journal, Vodafone, Wired

AI / ML Platform (12)

Anthropic, Claude, Cohere, ElevenLabs, MiniMax, Mistral AI, OpenAI, Perplexity, Runway, Together AI, VoltAgent, xAI

Productivity / Collaboration (8)

Airtable, Cal.com, Figma, Intercom, Linear, Miro, Notion, Zapier

Automotive (8)

BMW, BMW M, Bugatti, Ferrari, Lamborghini, Renault, SpaceX, Tesla

Why this matters

The other 8 popular Claude UX plugins generate "modern minimal" or "clean dashboard" — variants of the same default aesthetic. ux-skill lets you ask for Linear's clarity, Stripe's seriousness, Apple's restraint, Tesla's monolith, Notion's friendliness, Cursor's gradient discipline, Raycast's hairline density, Claude's warm editorial — and the engine pulls the right tokens, voice, motion conventions, and signature moves from the brand spec.

MCP server — the asymmetric move

ux-skill ships a Model Context Protocol server. Run ux-mcp and the engine becomes a long-running stdio process that any MCP-capable host — Claude Desktop, Cursor, Windsurf, generic agents — can call into. Eighteen tools: ux_recommend, ux_lint, ux_styles, ux_palettes, ux_type_pairs, ux_components, ux_industries, ux_motion_presets, ux_anti_patterns, ux_brands, ux_landing_patterns, ux_persist_save, ux_persist_load, ux_stats, ux_image_extract, ux_synthesize, ux_decisions_query, ux_decisions_stats. Same Python handlers the slash commands use; same data manifests; same deterministic recommender.

Why this is the asymmetric move: none of the top eight Claude UX skills (ui-ux-pro-max-skill, open-design, taste-skill, huashu-design, stitch, nothing-design, hallmark, material-3) ship an MCP server. They are locked inside Claude Code's plugin runtime. ux-skill is reachable from any host that speaks MCP, including agents that have never heard of a Claude Code plugin.

pip install 'uxskill[mcp]'             # mcp is an opt-in extra
ux-mcp                                  # stdio JSON-RPC server starts

Point your client at the ux-mcp binary. Full tool docs, JSON examples, and per-client config for Claude Desktop, Cursor, and Windsurf live at docs/mcp.html and in commands/ux-mcp.md.

The 17-IDE installer

uxskill init (or /ux-init inside Claude Code) auto-detects which IDE you're using and writes the right artifact. Same Python engine. Same recommendations. Different glue per IDE.

IDE / Tool	Detection signal	Installed artifact
Claude Code	`.claude/` or `CLAUDE.md`	Plugin manifest at `.claude-plugin/plugin.json` + all 25 commands + all 5 sub-agents
Cursor	`.cursor/` or `.cursorrules`	`.cursorrules` prompt header pointing at the engine
Windsurf	`.windsurf/` or `.windsurfrules`	`.windsurfrules` with the same prompt header
GitHub Copilot	`.github/copilot-instructions.md` or `.vscode/`	`.github/copilot-instructions.md`
Gemini CLI	`GEMINI.md`	`GEMINI.md`
Codex	`AGENTS.md`	`AGENTS.md`
Kiro	`.kiro/`	`.kiro/instructions.md`
Cline	`.cline/`	`.cline/instructions.md`
Continue	`.continue/`	`.continue/config.json` patch
Aider	`.aider.conf.yml`	`.aider.conf.yml` + `AIDER.md`
Zed	`.zed/`	`.zed/instructions.md`
JetBrains AI	`.jetbrains-ai/` or `.idea/`	`.jetbrains-ai/instructions.md`
Pieces	`.pieces/`	`.pieces/instructions.md`
Tabby	`.tabby/`	`.tabby/instructions.md`
Tabnine	`.tabnine/`	`.tabnine/instructions.md`
CodeWhisperer	`.aws-codewhisperer/`	`.aws-codewhisperer/instructions.md`
Roo Cline	`.roo/`	`.roo/instructions.md`

In every IDE, the same uxskill recommend / uxskill lint / uxskill stats CLI commands work from the terminal. The Python engine is the source of truth; the IDE artifacts are thin prompt-headers that route into it.

Use cases — concrete scenarios

Eight real scenarios. Pick the one closest to your situation and adapt the invocation.

1. Building a fintech dashboard in Cursor

You're in Cursor working on a MENA neobank dashboard. You install the plugin and run discovery, recommendation, then dashboard generation.

pip install uxskill
uxskill init                                # detects Cursor, writes .cursorrules
uxskill discover                            # 10-field intake
uxskill recommend \
  --project-type=dashboard \
  --industry=fintech-neobank \
  --tone=clinical --tone=precise \
  --must-have=dark-mode --must-have=a11y-AA --must-have=RTL \
  --forbidden=brutalism --forbidden=purple-gradients \
  --stack=nextjs-15-app-router \
  --region=mena

Then in Cursor, ask: "Generate the dashboard surface using the recommendation in .ux/last-recommendation.json". Cursor reads the .cursorrules header, loads the recommendation, dispatches a dashboard generation with explicit constraints.

2. Generating a Stripe-style landing in Claude Code

/plugin marketplace add Laith0003/ux-skill
/plugin install ux@ux-skill
/ux-discover
> Project type? landing
> Industry? fintech-payments
> Tone? serious, technical, confident
> Must have? dark-mode, AA, mobile-first
> Forbidden? purple-gradients, three-equal-cards
> Reference brands? stripe
> Stack? nextjs-15-app-router
> Region? global
> Success metric? signup conversion

/ux-recommend
> [returns picked style, palette, type pair, motion presets, components, brand exemplars]

/ux-design "generate the landing using the Stripe brand spec as exemplar"
> [frontend-engineer generates the page]

/ux-lint .
> [passes — Stripe brand spec was respected]

3. Auditing existing code for AI slop in CI

You shipped a Next.js app two weeks ago. You want a hard floor against AI fingerprints on every PR.

# .github/workflows/ux-lint.yml
name: ux-lint
on: [pull_request]
jobs:
  lint:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with: { python-version: '3.11' }
      - run: pip install uxskill
      - run: uxskill lint . --ci --fail-on high

PRs that introduce purple-to-blue gradients, Inter at 96px, "John Doe" testimonials, or emoji-as-icons fail CI. No LLM cost. ~200ms.

4. Polishing an existing surface that "feels AI-generated"

You inherited a React app that looks like every other AI-generated SaaS site. You want to make it not look like that.

/ux-critique src/components/Hero.tsx
> [3 wins, 3 misses, 1 strategic move — the take is honest]

/ux-lint src/
> [15 high-severity AI fingerprints flagged]

/ux-polish src/components/Hero.tsx
> [LLM-driven cosmetic pass + AI-slop kill]

/ux-fix
> [applies findings as atomic commits, re-runs the linter]

Three commands, one polished surface, atomic commits per fix.

5. Designing a Linear-style command palette

/ux-component command-palette --brief="Linear-style, dark, monospace shortcuts, recent items first"
> [reads data/brands/linear.app.json for tokens + signature moves]
> [reads data/components.json for the command-palette anatomy + states]
> [dispatches frontend-engineer with explicit Linear spec]

The generated component uses Linear's actual color tokens, type stack, motion conventions, hairline densities — not "generic dark UI."

6. Running a 90-minute design thinking workshop with stakeholders

You have a room of 5 people for 90 minutes. You want them to leave with a game plan, not a vibe.

/ux-workshop "loyalty wallet pivot" \
  --participants="2 PMs, 1 designer, 1 eng lead, 1 customer rep" \
  --minutes=90

The plugin facilitates the five phases (exploration → heat map → stakeholder map → solution sketch → game plan) end-to-end, time-boxed, with concrete per-phase artifacts. The output is .ux/last-workshop.json — the game plan, not just "interesting findings."

7. Writing a publishable case study after launch

You shipped the loyalty wallet. You want a portfolio piece.

/ux-case-study --format=html --slug=bashiti-loyalty
> [reads .ux/last-frame.json, last-workshop.json, last-research.json, last-design.json, last-a11y.json, last-polish.json, last-recommendation.json, last-discovery.json]
> [generates Wfrah-editorial case study with numbered (A)-(G) sections, hairline separators, bilingual-safe layout]
> [writes case-studies/bashiti-loyalty.html]

The case study is a finished, publishable artifact — not a draft. Pure monochrome, editorial typography, ready to ship to your portfolio.

8. Running discovery in a non-AI context (just structured intake)

You're scoping a project. You don't need a recommendation yet — you need a structured brief.

uxskill discover
# 10-field intake, saves to .ux/last-discovery.json

cat .ux/last-discovery.json
# {
#   "project_type": "...",
#   "audience": "...",
#   ...
# }

You can hand the JSON to your team, paste into a Notion doc, or feed it into a separate AI tool. ux-skill is also a structured intake tool in addition to being an engine.

9. MASTER.md persistence — your design decisions, in the repo

After /ux-recommend, persist the picked style + palette + type + motion + components + brand exemplars + guardrails as a human-readable Markdown file your team can review, diff, and version-control.

python3 -m engine.cli.main persist save --project-root .

Writes .ux/design-system/MASTER.md (YAML frontmatter + body) and .ux/design-system/pages/<name>.md per generated surface via persist save-page. Idempotent — same input produces byte-identical output, so re-running on unchanged state is a no-op in git.

Compared to alternatives

Short summary table. Full table-by-table comparison is at uxskill.laithjunaidy.com/compare.html.

Dimension	ux-skill	ui-ux-pro-max	open-design	taste-skill	huashu-design	stitch-skills	nothing-design	hallmark	material-3
Slash commands	22	1	19	1	1	multi	1	1	1
Components	148	0	0	0	0	0	0	0	(MD3)
Motion presets	57	0	0	0	0	0	0	0	0
Brand specs	160	0	0	0	0	0	0	0	0
Anti-pattern rules	145	0	0	0	0	0	0	0	0
CI-safe deterministic linter	yes	no	no	no	no	no	no	no	no
IDEs supported	17	18	1	1	1	1	1	1	1
Discovery gate	10 fields	implicit	implicit	implicit	implicit	implicit	implicit	implicit	implicit
`.ux/` state chain	yes	no	no	no	no	no	no	no	no
Stars (2026-05-28)	14	83,958	54,406	25,202	15,455	5,762	2,391	2,164	955

Honest assessment

ui-ux-pro-max is bigger on awareness, ships 18 IDEs, has BM25-style search over its CSV. It doesn't ship a component manifest, motion manifest, brand library, or deterministic linter.
open-design has 19 skills + preview but only Claude Code support and no anti-slop layer.
hallmark is the closest in spirit (also anti-slop) but is a single skill — no engine, no manifests, no chained commands.
material-3-skill is excellent if you specifically want Material Design 3. We don't compete on MD3.

For full detail per dimension, see compare.html.

Roadmap

v2.1 — Linter completeness (Q3 2026)

+17 deferred anti-pattern rules to reach 52 total. Targets: dark-on-dark hover states, color-only state encoding, redundant z-index escalation, hardcoded breakpoints in JS, opacity instead of disabled state, etc.
uxskill lint --fix for safe rewrites of mechanically-fixable findings (button-no-type, img-no-alt empty-string, console-log-leak removal).
VS Code extension that surfaces lint findings inline (no need to run CI).

v2.2 — Component manifest expansion (Q4 2026)

+50 components to reach 198 total. Net-new: combobox with async filter, command-palette with recent-items heuristic, conditional-form-step, payment-element variants, RTL-aware date picker, MENA-specific phone input, calendar grid with hijri overlay.
Per-component code emit in 6 stacks (Next.js + React, Vue 3 + Nuxt, SvelteKit, Astro, Blade + Alpine, vanilla HTML/CSS).
Component playground at uxskill.laithjunaidy.com/playground — try the recommendation engine + see live component preview.

v3 — The marketplace + the lock-in (2027)

Brand spec marketplace — publish and discover community brand specs. Pay-to-publish to fund moderation.
Custom anti-pattern rules — projects can define their own regex rules in data/anti-patterns.local.json (already shipped in v2; v3 adds discovery + sharing).
uxskill plan — full multi-page site planning from a brief, not just one surface.
Figma plugin parity — same recommendation engine, surfaced in Figma.

Contributing

Issues and PRs welcome. Three high-leverage areas:

Add an anti-pattern rule

Edit data/anti-patterns.json — add an entry with id, name, severity, category, detection.pattern, detection.flags, detection.scope, evidence_template, fix, references.
Add a test in tests/linter/ — one file that triggers the rule, one that doesn't.
Run uxskill lint tests/linter/should-trigger/<rule>.tsx — confirm it fires. Run on tests/linter/should-not-trigger/<rule>.tsx — confirm it doesn't.
Open a PR.

Add a brand spec

Create data/brands/<slug>.json with id, name, category, voice, tokens, design_principles, signature_moves, anti-moves, references.
Add the corresponding prose at references/brands/<slug>.md.
Register in data/brands/_index.json.
Open a PR. Spec must be backed by primary-source references (the brand's actual product, public design system, or DESIGN.md if they publish one).

Add a motion preset

Edit data/motion-presets.json — add an entry with id, name, category, tokens, stacks (framer_motion, gsap, css), accessibility.reduced_motion_fallback, when_to_use.
The preset must have a reduced-motion variant. No exceptions.
Open a PR.

Process

Read CONTRIBUTING.md for the full process.
Read CODE_OF_CONDUCT.md.
New rules and brand specs are reviewed for: primary-source grounding, no overfitting to a single project, no emoji in any of the data, RTL-safe behavior where applicable.

License, author, acknowledgments

License

MIT. Use it, fork it, build on it. If it saves you from shipping AI slop, star the repo — it's the cheapest way to support it.

Author

Laith Aljunaidy — solo founder of Dot, a MENA-first loyalty platform. Building ux-skill so the AI-generated frontend doesn't all look the same.

LinkedIn: linkedin.com/in/laithaljunaidy
Email: laith.aljunaidy.laith@gmail.com
Repo: github.com/Laith0003/ux-skill
Site: uxskill.laithjunaidy.com
PyPI: pypi.org/project/uxskill
npm: npmjs.com/package/uxskill

Acknowledgments

The team at Anthropic for Claude Code and the skill / plugin architecture that made this distributable.
Nielsen Norman Group, Laws of UX (lawsofux.com), and the UX research community whose work informs data/ux-guidelines.json.
Every brand listed in data/brands/ — their public design systems are the source of truth for the brand specs.
The original v1 contributors: a single-shot Claude skill that became the seed for the v2 Python engine.
The 8 popular Claude UX plugins we compared against — they raised the bar; this is our answer.

ux-skill · v3.0.0-stable · Built so Claude Code, Cursor, Windsurf, and every other AI coding tool output frontend that doesn't read as AI-generated.

Star the repo at github.com/Laith0003/ux-skill · Install via pip install uxskill or npx uxskill init · Browse the comparison at uxskill.laithjunaidy.com/compare.html

在 Playground 中尝试

服务器配置

{
  "mcpServers": {
    "ux-skill": {
      "command": "ux-mcp"
    }
  }
}

Build with ShipAny.

Ux Skill

ux-skill — the design intelligence engine for Claude Code, Cursor, and every other AI coding tool

New in v3.1 — brand-true, responsive, alive

What's new in v3

Star history

What is ux-skill

Table of contents

The Brain — what v3.0 is

Quick install

Path 1 — Claude Code marketplace (canonical)

Path 2 — pip (universal)

Path 3 — npx (no Python required)

Verify install

The numbers — live comparison vs the top 8 Claude UX skills

Where we lose

Where we win

Architecture — how the pieces fit

How the engine actually works

The 25 slash commands — detailed reference

Bootstrap & inventory

/ux-init — bootstrap the project

/ux-stats — print the data inventory

Discovery & recommendation

/ux-discover — the forcing function (10-field intake)

/ux-recommend — the flagship 5-parallel-search engine

Generation

/ux-design — generate a beautiful, anti-slop surface from a brief

/ux-component — generate one component

/ux-system — generate a complete starter design system

/ux-dashboard — specialized dashboard generation

/ux-motion — motion treatment

Audit & verify

/ux-lint — deterministic regex-based linter (no LLM, CI-safe)

/ux-audit — 6-lens design audit

/ux-a11y — WCAG 2.1 AA audit + common-courtesy checks

/ux-critique — taste call (3 wins, 3 misses, 1 strategic move)

/ux-copy — microcopy review + rewrite

Fix & polish

/ux-fix — apply findings as atomic commits

/ux-polish — cosmetic pass + AI-slop kill