home / skills / openclaw / skills / ui-ux-pro-max
This skill provides UI/UX design guidance and concrete frontend-ready specifications to craft polished interfaces across web and mobile.
npx playbooks add skill openclaw/skills --skill ui-ux-pro-maxReview the files below or copy the command above to add this skill to your agents.
---
name: ui-ux-pro-max
description: UI/UX design intelligence and implementation guidance for building polished interfaces. Use when the user asks for UI design, UX flows, information architecture, visual style direction, design systems/tokens, component specs, copy/microcopy, accessibility, or to generate/critique/refine frontend UI (HTML/CSS/JS, React, Next.js, Vue, Svelte, Tailwind). Includes workflows for (1) generating new UI layouts and styling, (2) improving existing UI/UX, (3) producing design-system tokens and component guidelines, and (4) turning UX recommendations into concrete code changes.
---
Follow these steps to deliver high-quality UI/UX output with minimal back-and-forth.
## 1) Triage
Ask only what you must to avoid wrong work:
- Target platform: web / iOS / Android / desktop
- Stack (if code changes): React/Next/Vue/Svelte, CSS/Tailwind, component library
- Goal and constraints: conversion, speed, brand vibe, accessibility level (WCAG AA?)
- What you have: screenshot, Figma, repo, URL, user journey
If the user says "全部都要" (design + UX + code + design system), treat it as four deliverables and ship in that order.
## 2) Produce Deliverables (pick what fits)
Always be concrete: name components, states, spacing, typography, and interactions.
- **UI concept + layout**: Provide a clear visual direction, grid, typography, color system, key screens/sections.
- **UX flow**: Map the user journey, critical paths, error/empty/loading states, edge cases.
- **Design system**: Tokens (color/typography/spacing/radius/shadow), component rules, accessibility notes.
- **Implementation plan**: Exact file-level edits, component breakdown, and acceptance criteria.
## 3) Use Bundled Assets
This skill bundles data you can cite for inspiration/standards.
- **Design intelligence data**: Read from `skills/ui-ux-pro-max/assets/data/` when you need palettes, patterns, or UI/UX heuristics.
- **Upstream reference**: If you need more phrasing/examples, consult `skills/ui-ux-pro-max/references/upstream-skill-content.md`.
## 4) Optional Script (Design System Generator)
If you need to quickly generate tokens and page-specific overrides, use the bundled script:
```bash
python3 skills/ui-ux-pro-max/scripts/design_system.py --help
```
Prefer running it when the user wants a structured token output (ASCII-friendly).
## Output Standards
- Default to ASCII-only tokens/variables unless the project already uses Unicode.
- Include: spacing scale, type scale, 2-3 font pair options, color tokens, component states.
- Always cover: empty/loading/error, keyboard navigation, focus states, contrast.
This skill provides UI/UX design intelligence and practical implementation guidance for building polished interfaces across web and native platforms. It helps you create or improve layouts, design systems, component specs, microcopy, accessibility, and front-end code (HTML/CSS/JS, React, Next.js, Vue, Svelte, Tailwind). Use it to move from high-level UX recommendations to concrete, testable design and implementation artifacts.
I start with a focused triage to collect only essential constraints: platform, tech stack, goals, and artifacts you already have (screenshots, Figma, repo, URL). Then I produce concrete deliverables: UI concept and layout, UX flow mapping, design-system tokens and component rules, and a file-level implementation plan with acceptance criteria. When requested, I can generate ASCII-friendly token output using a bundled script for fast, structured token lists.
What do you need to start quickly?
Provide platform, stack, primary goal, and one artifact (screenshot, Figma link, or repo). I only ask essentials to avoid wasted work.
Can you produce code I can drop into my repo?
Yes. I deliver file-level edits, component specs, and explicit acceptance criteria. For token lists I default to ASCII-friendly output and can run the generator script when requested.