agentleFS
mike22890mike22890/opencode-skills-en/skills/applying-ui-design-system/SKILL.md

applying-ui-design-system

Skill1 starsChanged 2 months agoMITView on GitHub

前端 UI 设计系统的选型与实现。当需要为产品选择视觉风格(Glassmorphism、Liquid Glass、Neubrutalism、Claymorphism、Neo-brutalism、拟物、扁平、极简、暗黑模式)、实现设计 token、CSS 变量、组件样式、主题切换、设计系统搭建、Tailwind 配置、shadcn/ui 定制时使用。

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

设计语言选型

何时触发

关键词命中即触发:设计系统/风格/主题/暗黑/玻璃/拟物/新拟物/token/组件/Tailwind/shadcn。

判断先于执行:决定「是否做 / 选什么 / 如何取舍」(栈、方案、架构、权衡)前,先读领域判断内核 _kernel-frontend/README.md——它管 judgment,本秘典管 execution;冲突时以内核判断为准。

不同设计语言不可混搭。先选语言,再按 reference 落地组件。

风格决策

想要的感觉选加载
半透明、磨砂、轻盈、纵深感弱Glassmorphismreferences/glassmorphism.md
Apple 风格、立体玻璃、环境光响应、弹簧动效Liquid Glassreferences/liquid-glass.md
粗框硬影、高饱和、纯平、反精致Neubrutalismreferences/neubrutalism.md
软糯黏土、大圆角、双内影、温暖Claymorphismreferences/claymorphism.md

每个 reference 自包含「核心要素 + CSS Tokens + 关键约束 + 组件 + 暗色模式 + 无障碍」全套。

共享铁律(跨风格)

无论选哪种风格,下列约束全部适用:

  1. 对比度 — 主文字 ≥ 4.5:1,大字体(≥18pt 常规 / ≥14pt 粗体)≥ 3:1。逐底色验证,不靠白底脑补。
  2. :focus-visible — 必须可见;轮廓与底色有足够反差。键盘可达性是底线。
  3. prefers-reduced-motion: reduce — 禁用弹簧、滑入、parallax 等装饰动画;保留功能性 transition。
  4. prefers-contrast: more — 减少透明、增强边界、平化阴影。
  5. 暗色模式 — 用语义 token 切换(如 --bg-surface),不要硬编码 #fff/#000 后再交换。

落地流程

  1. 选风格 — 一个项目通常一种主风格;可有第二风格作辅助层(如主体 Liquid Glass + 强调按钮 Neubrutalism)
  2. 导入 tokens — 复制对应 tokens-<style>.css 到项目,按需扩展
  3. 按 reference 抄组件 — Card / Button / Modal / Nav 速查已就位
  4. 验证对比度 + 焦点 + 减弱动效 — 跑 a11y 自检
  5. 暗色模式映射 — 每个 token 都要有暗色对应

何时不要本 skill

  • 仅需 vanilla CSS / Tailwind 默认组件 → 不需要设计系统选型
  • 已有完整 design system(如 Material、Ant Design)→ 沿用,不重新选风格
  • 后端工作 / 数据可视化主导 → 走 designing-architectures

性能基线

  • 同屏 backdrop-filter 元素 ≤ 3(Glassmorphism / Liquid Glass 限制)
  • 滚动容器内避免 blur(触发逐帧重绘)
  • 阴影层叠 ≤ 3 层(Claymorphism 多内影时尤其注意)
  • 低端设备:用 @supports 或 @media (hover: none) 降级为纯色

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.