Token导航 LogoToken导航TokenDH.com
研究检索external-servicegithub未标认证来源可访问许可证需确认审计通过

visual-hierarchy视觉层次

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

326

周安装

14

GitHub Stars

12

下载量

114
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:visual-hierarchy(视觉层次)
来源仓库:https://github.com/claude-dev-suite/claude-dev-suite
仓库路径:skills/visual-hierarchy
安装命令:
npx skills add https://github.com/claude-dev-suite/claude-dev-suite --skill visual-hierarchy
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/claude-dev-suite/claude-dev-suite --skill visual-hierarchy

简介

用于界面设计的视觉层次与排版规范指导。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

  • 适合优化字体缩放、间距一致性和信息优先级展示。
  • 使用时需结合品牌指南和用户任务流组织内容结构。
  • 涉及实际页面时应通过截图检查文本截断与对齐问题。
  • visual-hierarchy 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Visual Hierarchy

Fluid Typography Scale

Use clamp(min, preferred, max) for fluid type that scales between breakpoints without jumps.

:root {
  /* Fluid type scale — adapts between ~375px and ~1280px viewport */
  --text-xs:   clamp(0.75rem,  0.70rem + 0.25vw, 0.875rem);
  --text-sm:   clamp(0.875rem, 0.82rem + 0.30vw, 1rem);
  --text-base: clamp(1rem,     0.92rem + 0.40vw, 1.125rem);
  --text-lg:   clamp(1.125rem, 1.00rem + 0.60vw, 1.375rem);
  --text-xl:   clamp(1.25rem,  1.10rem + 0.75vw, 1.75rem);
  --text-2xl:  clamp(1.5rem,   1.25rem + 1.25vw, 2.25rem);
  --text-3xl:  clamp(1.875rem, 1.50rem + 1.90vw, 3rem);
  --text-4xl:  clamp(2.25rem,  1.75rem + 2.50vw, 3.75rem);
}

Key rules:

  • Body text: 16px minimum (1rem) — never smaller for continuous reading
  • Line height: 1.4–1.6 for body text; 1.1–1.3 for headings
  • Optimal line length: 45–90 characters desktop, 30–50 characters mobile
  • Letter spacing: 0 to +0.02em for body; −0.02em to −0.04em for large headings

Spacing System (4px base grid)

All spacing values should be multiples of 4px for visual consistency.

:root {
  --space-1:  0.25rem;  /*  4px */
  --space-2:  0.5rem;   /*  8px */
  --space-3:  0.75rem;  /* 12px */
  --space-4:  1rem;     /* 16px */
  --space-5:  1.25rem;  /* 20px */
  --space-6:  1.5rem;   /* 24px */
  --space-8:  2rem;     /* 32px */
  --space-10: 2.5rem;   /* 40px */
  --space-12: 3rem;     /* 48px */
  --space-16: 4rem;     /* 64px */
  --space-20: 5rem;     /* 80px */
  --space-24: 6rem;     /* 96px */
}

User Scanning Patterns

Research from Nielsen Norman Group eye-tracking studies (2006–2024):

PatternShapeWhen it occursLayout implication
F-patternTwo horizontal bands + left verticalText-heavy pages, unfocused readingPlace key info top-left; front-load headings with meaningful words
Z-patternDiagonal from top-left to bottom-rightLanding pages, minimal UILogo top-left, CTA bottom-right, key message along diagonal
Layer-CakeHorizontal bands (headings only)Motivated users scanning for infoUse meaningful subheadings; bold keywords; avoid generic labels
SpottedJump to visuals/links/bold textUsers seeking specific contentUse visual anchors; highlight action words; bold CTAs
CommitmentTop-to-bottom linearHigh-motivation, interested usersReward with well-structured content; don't waste early paragraphs

Practical rules:

  • First 2 words of a heading carry 80% of scanning weight — make them meaningful
  • Users read ~28% of words on a page on average
  • Layer-Cake is the most effective pattern to design for after initial F-pattern entry

Cognitive Load Reduction

Miller's Law: Working memory holds 7 ± 2 chunks. Keep navigation items, form fields, and choice options within this range.

Progressive Disclosure types:

  1. Staged disclosure — linear steps (wizard, onboarding)
  2. Contextual disclosure — show details relevant to current action (tooltip, expand)
  3. Feature-based — reveal advanced options only when needed ("Advanced settings...")
  4. Interaction-triggered — appear on demand (accordion, "Show more")

Chunking patterns:

Bad:  12 navigation items in a flat list
Good: 4 groups × 3 items each (same 12 items, 4 cognitive units)

Bad:  Phone: 0612345678
Good: Phone: 06 1234 5678  (chunked into familiar groups)

Hick's Law: Decision time increases logarithmically with number of choices. Reduce choices at every decision point.


Visual Weight Principles

Visual weight determines what users look at first. In descending order of weight:

FactorHigh weightLow weight
SizeLarge elementsSmall elements
ContrastHigh contrast vs backgroundLow contrast
ColorSaturated / warm colorsDesaturated / cool / gray
PositionTop-left, centerBottom-right, edges
IsolationSurrounded by whitespaceCrowded with other elements
MotionAnimated elementsStatic elements
ShapeIrregular, complexRectangular, simple

Hierarchy creation rule: For any screen, one element should have clearly the highest visual weight (primary focus), one or two secondary, and the rest tertiary. Multiple "primary" elements = no hierarchy.


Whitespace as Design Element

Whitespace is not empty — it actively communicates grouping, importance, and breathing room.

Gestalt proximity: Elements close together are perceived as a group. Use consistent internal vs. external spacing.

Component internal spacing: --space-3 to --space-4  (12–16px)
Between components:         --space-6 to --space-8  (24–32px)
Section separations:        --space-12 to --space-16 (48–64px)
Page padding (mobile):      --space-4  (16px min)
Page padding (desktop):     --space-8 to --space-16

Macro vs micro whitespace:

  • Micro: Padding inside buttons, cells, inputs — affects scanability
  • Macro: Space between sections and components — affects perceived quality; generous macro whitespace signals premium

Research finding: Increasing whitespace around text and titles increases comprehension by 20%.


Related Skills

  • styling/tailwindcss — implementation utilities (spacing, text sizing)
  • styling/shadcn-ui — component-level whitespace and typography tokens
  • accessibility/wcag — contrast ratios for color in hierarchy
  • ux/design-systems — tokens that codify the spacing and type scale
  • ux/interaction-design — motion and loading states

Deep Knowledge

Load via mcp__documentation__fetch_docs:

  • ux-visual-hierarchy — full scanning pattern data, cognitive load research, fluid typography reference

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

36.16%
按下载量换算41

Claude

29.47%
按下载量换算34

Cursor

19.26%
按下载量换算22

Gemini CLI

9.52%
按下载量换算11

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

external-service

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

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