Token导航 LogoToken导航TokenDH.com
前端设计操作浏览器github未标认证来源可访问许可证需确认审计通过

ui-design用户界面设计

Agent Skill

用于辅助界面设计、视觉规范、排版、配色、布局和交互体验优化。它适合让 Agent 根据产品场景整理页面结构、生成 UI 方案、检查视觉一致性或改进组件层级。使用时需要结合现有品牌、设计系统和用户任务,不应只堆装饰元素;涉及真实页面改动时,应通过截图或浏览器预览检查文本溢出、对齐和响应式表现。

总安装

392

周安装

16

GitHub Stars

6

下载量

127
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:ui-design(用户界面设计)
来源仓库:https://github.com/sebnow/configs
仓库路径:skills/ui-design
安装命令:
npx skills add https://github.com/sebnow/configs --skill ui-design
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/sebnow/configs --skill ui-design

简介

ui-design 用于辅助界面设计、视觉规范和交互体验优化。

  • 适合整理页面结构、生成 UI 方案或检查视觉一致性。
  • 使用时需结合品牌和设计系统,避免堆砌装饰元素。
  • 涉及真实页面改动时应通过截图或浏览器预览检查文本溢出和对齐。
  • 适用于 Codex、Claude、Cursor、Gemini CLI 等宿主环境。

SKILL.md

UI Design

Required when implementing visual design and styling interfaces. Enforces intentional aesthetics and systematic implementation to avoid generic patterns.

Core Principles

You must follow these non-negotiable principles:

  1. Intentionality over intensity - Bold maximalism and refined minimalism both work; generic mediocrity fails
  2. Design tokens are mandatory - Systematic values prevent inconsistency
  3. Avoid AI slop aesthetics - Generic patterns destroy brand identity
  4. Implementation matches vision - Maximalist designs require elaborate code; minimalist designs demand precision
  5. AI personalization requires ethics - Transparency and user control are non-negotiable

Mandatory Pre-Implementation

Never start coding without defining:

Aesthetic Direction (Required)

Commit to a bold aesthetic choice before implementation:

Tone Options:

  • Brutalist (raw, functional, high contrast)
  • Maximalist (layered, rich, abundant detail)
  • Retro-futuristic (nostalgic tech, bold geometry)
  • Organic (natural forms, flowing lines)
  • Luxury (refined, elegant, spacious)
  • Playful (unexpected, whimsical)
  • Editorial (typographic hierarchy, grid-based)
  • Art deco (geometric glamour, bold symmetry)
  • Industrial (mechanical, utilitarian)
  • Soft (gentle curves, muted tones)

State your chosen aesthetic and commit to it throughout implementation.

Generic "clean" and "modern" are not acceptable design directions.

Technical Context

Document before proceeding:

  1. Framework/library constraints
  2. Performance budget
  3. Browser/device support requirements
  4. Existing design system (if any)
  5. AI personalization needs

Design Token System (Non-Negotiable)

You must implement systematic design tokens. Never use magic numbers or arbitrary values.

Token Structure

Use three-tier system:

  1. Foundation tokens: Raw values (#3b82f6, 0.25rem)
  2. Semantic tokens: Purpose-driven (--color-primary, --space-base)
  3. Component tokens: Specific usage (--button-color-primary-bg)

Naming Convention

Use: category-property-variant-state structure.

Examples: --color-text-body, --button-color-primary-hover, --space-stack-large

Required Categories

  • Color: primary, secondary, accent, text, background, surface, border, states
  • Spacing: 8px base system (space-1 through space-24)
  • Typography: Consistent scale (1.25 or 1.414 ratio)
  • Border radius: sm, md, lg, xl, full
  • Shadows: sm, md, lg, xl, 2xl

See @tokens.md for detailed token definitions and platform-specific generation.

Avoiding Generic AI Aesthetics

These patterns signal low-quality AI-generated design.

Forbidden Patterns

Typography: Inter/Roboto without justification, single font weight, uniform sizing Color: Purple-blue gradients by default, pastel without contrast, gray-on-gray-on-gray Layout: Everything centered equally, only card grids, no asymmetry, uniform spacing Motion: Decorative animations, simultaneous animations, fade-in only, no easing

Required Distinctiveness

Typography: Distinctive fonts with personality, contrasting weights (300/400 body, 600/700/800 headings), variable fonts Color: Dominant color + sharp accents, intentional contrast levels, documented usage Composition: Asymmetry, overlap for depth, diagonal flow, grid-breaking moments, negative space as design element

Frontend Implementation Guidelines

Typography Implementation

Use distinctive fonts with proper hierarchy. Never default to Inter/Roboto without justification.

Required elements:

  • Font pairing (display + body)
  • Weight hierarchy (300/400 for body, 600/700/800 for headings)
  • Variable fonts for fluid control
  • Proper line-height and letter-spacing

See @typography.md for implementation patterns and font loading.

Color & Theme Implementation

Use CSS custom properties for theme switching. Map semantic tokens to foundation tokens per theme.

Required contrast ratios:

  • Text: 4.5:1 minimum (WCAG AA)
  • Interactive components: 3:1 minimum

See @tokens.md for dark mode implementation patterns.

Motion & Animation

Prioritize high-impact moments over scattered effects.

Required:

  • Proper easing curves (never linear)
  • Respect prefers-reduced-motion
  • Performance-conscious implementation
  • Purposeful animations only

See @animation.md for orchestration patterns and scroll-triggered effects.

Background & Atmosphere

Create visual richness through layering when aesthetic requires it. See @backgrounds.md for gradient meshes, textures, and patterns.

Component Library Patterns

Each component must:

  • Define its own tokens
  • Document all interactive states (default, hover, active, focus, disabled)
  • Use semantic HTML
  • Include keyboard focus indicators

See @components.md for implementation patterns and design system tools.

AI-Driven Personalization

When implementing AI-powered interface adaptation:

Ethical Requirements (Non-Negotiable)

  1. Transparency: Users must know when AI personalizes their experience
  2. Control: Provide opt-out and reset options
  3. Explanation: Show why adaptations occurred
  4. Human fallback: Offer support when AI fails
  5. Privacy: Never personalize with sensitive data without explicit consent

Personalization Boundaries

Appropriate:

  • Layout density preferences
  • Color scheme (light/dark mode)
  • Content prioritization
  • Navigation shortcuts

Forbidden:

  • Hiding critical functionality
  • Removing accessibility features
  • Changing core workflows without notice
  • Personalizing based on protected attributes

See @personalization.md for implementation patterns.

Implementation Checklist

Before marking UI design complete:

  • Aesthetic direction defined and documented
  • Design tokens implemented systematically
  • All token categories defined (color, spacing, typography)
  • No magic numbers in styles
  • Typography uses distinctive fonts with weight hierarchy
  • Color system is cohesive and intentional
  • Motion uses proper easing curves
  • prefers-reduced-motion respected
  • Component states documented (default, hover, active, focus, disabled)
  • Keyboard focus indicators visible
  • Design avoids generic AI patterns
  • AI personalization (if used) includes transparency and control
  • Visual hierarchy is clear
  • Spacing system is consistent

Red Flags

Stop if you catch yourself:

  • Using generic fonts (Inter/Roboto) or purple gradients without justification
  • Single font weight, uniform spacing, everything centered
  • Magic numbers instead of tokens
  • Animating everything simultaneously, skipping component states
  • AI personalization without consent UI
  • Copying patterns without adaptation

Deliverables

State: "UI design implementation complete" only after checklist passes.

Document for handoff:

  1. Design token definitions
  2. Component inventory with all states
  3. Aesthetic direction rationale
  4. Typography system (fonts, weights, scales)
  5. Color system (palette, usage, contrast ratios)
  6. Motion principles (easing, duration, triggers)
  7. AI personalization boundaries (if applicable)

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

38.93%
按下载量换算49

Claude

28.96%
按下载量换算37

Cursor

18.39%
按下载量换算23

Gemini CLI

10.36%
按下载量换算13

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills