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

design-critique设计批评

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

1,093

周安装

46

GitHub Stars

35

下载量

383
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:design-critique(设计批评)
来源仓库:https://github.com/petekp/claude-code-setup
仓库路径:skills/design-critique
安装命令:
npx skills add https://github.com/petekp/claude-code-setup --skill design-critique
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/petekp/claude-code-setup --skill design-critique

简介

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 方案或改进组件层级。
  • 需要结合品牌、设计系统和用户任务使用。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。
  • 涉及真实页面改动时应通过截图检查文本溢出和对齐。
  • design-critique 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Design Critique

Honest, specific, actionable feedback on interface design.

The Critique Stance

  • Be direct. No soft language. No vibes-only feedback.
  • Point to specifics, not generalities.
  • Explain why, not just what.
  • Reference principles, not preferences.
  • Offer fixes, not just problems.

Process

  1. Identify what you're critiquing - Screen, component, flow, or interaction
  2. Gather context - Platform, users, constraints (if available)
  3. Apply the lens - Clarity, hierarchy, interaction, accessibility, craft
  4. Prioritize issues - P0 (blocking) through P3 (polish)
  5. Propose fixes - Specific, actionable changes

Quick Check (Use CHECKLIST.md for detailed pass)

Clarity

  • Can users predict outcomes before acting?
  • Is the hierarchy of information obvious?
  • Are interactive elements clearly distinguished?

Interaction

  • Do all states exist? (hover, focus, active, disabled, loading, error)
  • Is feedback immediate?
  • Can users recover from errors?

Accessibility

  • Focus visible and logical?
  • Contrast sufficient?
  • Touch targets adequate (44x44pt)?

Output Contract

Structure every critique as:

## Verdict
[1 sentence: ship/iterate/rethink]

## Issues

### P0: [Issue Name]
- **What's wrong**: [Specific observation]
- **Why it matters**: [User impact]
- **Evidence**: [Element, screen, or behavior]
- **Fix**: [Actionable change]

### P1: [Issue Name]
...

## Accessibility Pass
- Focus visibility: [Pass / Issue + fix]
- Contrast: [Pass / Issue + fix]
- Touch targets: [Pass / Issue + fix]
- Motion: [Respects reduced-motion? / Issue]

## What's Working
[2-3 things done well - critique includes praise]

## Next Steps
- [ ] [Verification action 1]
- [ ] [Verification action 2]

Severity Levels

LevelDescriptionAction
P0Blocks task, causes confusion, or data lossMust fix before ship
P1Frequent friction, misclicks, unclear recoveryShould fix
P2Polish, efficiency, minor annoyanceFix if time permits
P3Nice-to-have refinementConsider for later

The Questions Behind Everything

  • "What is the user trying to do here?"
  • "What's the most important thing on this screen?"
  • "What would happen if we removed this?"
  • "Would a new user understand this?"
  • "Are we proud of this?"

Common Critique Notes

"Too busy": Too many things competing for attention. Remove until important things breathe.

"Not discoverable": Hidden functionality, unlabeled icons, gestures without affordance.

"Inconsistent": Different patterns for similar actions. Pick one and commit.

"Feels off": Usually spacing, alignment, or timing. The eye knows before the mind articulates.

"Overdesigned": Every effect turned up. Decoration overwhelming function. Subtract until inevitable.

Deep Reference

For detailed heuristics, common failures, and platform-specific patterns:

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

35.26%
按下载量换算135

Claude

34.22%
按下载量换算131

Cursor

18.16%
按下载量换算70

Gemini CLI

10.35%
按下载量换算40

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

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