Token导航 LogoToken导航

Agent Skill · 研究检索

steve-krug-ux-audit(史蒂夫·克鲁格 UX 审计)

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。

steve-krug-ux-audit 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为操作浏览器;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 20,093。

总安装 5,337周安装 218GitHub Stars 公开资料未说明下载 1,727核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.1

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:steve-krug-ux-audit(史蒂夫·克鲁格 UX 审计)
来源仓库:https://github.com/couriard/steve-krug-ux-audit
安装命令:
openclaw skills install steve-krug-ux-audit
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 OpenClaw 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。

ClawHubOpenClaw
openclaw skills install steve-krug-ux-audit

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

使用 Steve Krug 方法对网站或 Web 应用执行用户体验审核。提供结构化可用性审查依据,覆盖导航、文案与交互逻辑。输入 URL 或本地路径即可获得诊断报告与改进建议。适用于产品上线前评估或迭代优化阶段的质量把控。steve-krug-ux-audit 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

注意报告基于通用原则,需结合目标用户群体具体分析。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install steve-krug-ux-audit;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要操作浏览器,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

steve-krug-ux-audit 怎么安装?

在终端执行 openclaw skills install steve-krug-ux-audit。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

steve-krug-ux-audit 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

steve-krug-ux-audit 需要什么权限?

权限级别是操作浏览器。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。

steve-krug-ux-audit 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

steve-krug-ux-audit 用的人多吗?

公开数据是累计安装 5,337 次,近一周安装 218 次,下载 1,727 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

steve-krug-ux-audit 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
steve-krug-ux-audit

史蒂夫·克鲁格 UX 审计 · 本页

5,337—操作浏览器MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
steve-krug-ux-audit
description
Perform a UX audit on any website or web app using Steve Krug's proven methodology. Provide a URL or local path and get a structured usability review grounded in Krug's core principles, laws, and heuristics. Identifies issues from most critical to quick wins.
author
Chris Couriard
version
1.0.0
license
MIT
tags
[ux, usability, audit, design, web, accessibility, steve-krug]

Steve Krug UX Audit

Methodology Attribution

This skill applies Steve Krug's UX audit methodology — a proven approach to evaluating web usability and user-centered design. All core laws, heuristics, and frameworks referenced herein are drawn from Krug's established methodology.

Overview

Perform a practical usability audit of a website or web app using the principles from Steve Krug's *Don't Make Me Think, Revisited* (3rd Edition, 2014).

This skill evaluates real pages — via browser snapshots, screenshots, or provided URLs — against Krug's core laws, heuristics, and conventions. The output is an actionable audit report, not abstract theory.

When to Use

  • When provided with a URL or local dev path to review
  • When evaluating a page design, wireframe, or prototype for usability
  • When doing a pre-launch usability check
  • When reviewing competitor sites for UX comparison

How to Run the Audit

Step 1: Capture the Page

Use the browser tool to navigate to the URL/path and take a snapshot + screenshot. If multiple pages are specified, capture each one. At minimum, always audit:

  • The Home page
  • One key task flow (e.g. signup, purchase, search → result)

Step 2: The Trunk Test (Chapter 6)

Imagine being dropped on this page blindfolded. Can you answer these instantly?

QuestionWhat to Look For
What site is this?Site ID/logo — top-left, distinctive, recognizable at any size
What page am I on?Page name — prominent, in the right place, matches what you clicked
What are the major sections?Primary navigation — visible, consistent, well-labeled
What are my options at this level?Local/secondary navigation — clear subsections
Where am I in the scheme of things?"You are here" indicators — highlighted nav items, breadcrumbs
How can I search?Search box — visible, uses standard pattern (box + button + "Search")

Scoring: For each question, rate as ✅ Clear, ⚠️ Unclear, or ❌ Missing/Broken.

Step 3: Krug's Three Laws

First Law: "Don't make me think!" (Chapter 1)

Scan the page for question marks — anything that makes users pause and wonder:

  • Unclear names: Cute, clever, marketing-speak, or jargon labels instead of obvious ones
  • Ambiguous clickability: Links/buttons that don't look clickable, or non-links that do
  • Confusing choices: Options that require thought to distinguish (e.g. "RFPs" vs "Fixed-Price")
  • Unnecessary mental chatter: Form fields, labels, or flows that make users ask "What do they mean by...?"

Standard: The page should be self-evident (no thought required) or at minimum self-explanatory (a tiny amount of thought). If neither, it fails this law.

Second Law: "It doesn't matter how many times I have to click, as long as each click is a mindless, unambiguous choice." (Chapter 4)

Check all navigation paths and choices:

  • Are choices mindless (obvious what you'll get) or do they require thought?
  • Is the scent of information strong? (Do links clearly indicate their target?)
  • When choices are hard, is there just-in-time guidance (brief, timely, unavoidable)?
Third Law: "Get rid of half the words on each page, then get rid of half of what's left." (Chapter 5)

Evaluate text content:

  • Happy talk: Introductory/welcome text that says nothing useful ("Welcome to our amazing...")
  • Instructions nobody reads: Long instruction blocks instead of self-explanatory design
  • Needless words: Paragraphs that could be cut in half without losing meaning
  • Wall of words: Long unbroken paragraphs instead of short ones
  • Missing bullet lists: Series of items in paragraph form that should be bulleted

Step 4: Billboard Design (Chapter 3)

Users scan, they don't read. Evaluate the page as a billboard going by at 60 mph:

PrincipleCheck
ConventionsDoes it follow standard web conventions for layout, navigation, and element appearance?
Visual hierarchyAre important things more prominent? Are related things grouped? Is nesting clear?
Clearly defined areasCan you play "$25,000 Pyramid" — point at areas and say what they are?
Obvious clickabilityCan you instantly tell what's clickable and what's not?
Noise levelIs there shouting (everything competing)? Disorganization (no grid)? Clutter (too much stuff)?
Scannable textPlenty of headings? Short paragraphs? Bullet lists? Key terms bolded?
Heading hierarchyObvious visual distinction between heading levels? Headings closer to their section than the previous one?

Step 5: Navigation Design (Chapter 6)

Persistent Navigation
  • Present on every page (except forms)?
  • Contains: Site ID, Sections, Utilities, Search?
  • Consistent location and appearance throughout?
Page Names
  • Every page has one?
  • Prominent and in the right place (framing the content)?
  • Matches what user clicked to get there?
"You Are Here" Indicators
  • Current location highlighted in navigation?
  • Stands out enough? (Multiple visual distinctions — not too subtle)
Breadcrumbs (if applicable)
  • At the top of the page?
  • Uses ">" between levels?
  • Last item is bold and not a link?
Tabs (if used)
  • Active tab visually connects with content below?
  • Different color/shade from inactive tabs?

Step 6: Home Page (Chapter 7)

The Big Bang Theory of Web Design — first impressions are critical:

The Four Questions (must answer at a glance)
  1. What is this? — Is the site's purpose immediately clear?
  2. What can I do here? — Are available actions/content obvious?
  3. What do they have here? — Is the offering visible?
  4. Why should I be here and not somewhere else? — Is differentiation clear?
Key Elements
  • Tagline: Clear, informative, 6-8 words, conveys differentiation? (Not a vague motto)
  • Welcome blurb: Terse, prominent description of the site?
  • "Learn more" option: Video or explainer for complex propositions?
  • Entry points: Clear "where to start" for search, browse, and key tasks?
  • Promotional balance: Are promos overwhelming the core purpose (tragedy of the commons)?
The Fifth Question
  • Where do I start? — Is it obvious where to begin for search, browse, sign in, or the primary task?

Step 7: How Users Actually Behave (Chapter 2)

Evaluate the page against the three facts of web use:

  1. We don't read, we scan. Does the page support scanning? Or does it assume careful reading?
  2. We don't make optimal choices, we satisfice. Will the first reasonable-looking option lead users right? Or will satisficing lead them astray?
  3. We don't figure out how things work, we muddle through. Can someone muddle through successfully? Or will wrong mental models cause real problems?

Step 8: Mobile & Responsive (Chapter 10)

If reviewing mobile or responsive versions:

  • Prioritization: Are the most-needed features close at hand? Everything else reachable?
  • Affordances visible: Are tap targets obvious? (No hidden gestures, no reliance on hover)
  • Flat design tradeoffs: Has visual flattening removed useful affordances?
  • Performance: Any signs of bloat that would hurt mobile load times?
  • Zooming allowed? Can users zoom if they need to?
  • Deep links work? Do shared links go to the right page, not the home page?
  • Full site option? Is there a way to access the desktop version?

Step 9: Usability as Common Courtesy (Chapter 11)

Goodwill Drains (things that deplete trust)
  • [ ] Hiding information users want (pricing, support phone, shipping costs)
  • [ ] Punishing users for "wrong" formatting (phone numbers, credit cards)
  • [ ] Asking for unnecessary personal information
  • [ ] Faux sincerity / "shucking and jiving"
  • [ ] Sizzle blocking the path (marketing photos in the way of content)
  • [ ] Amateurish appearance
Goodwill Builders (things that increase trust)
  • [ ] Main user tasks are obvious and easy
  • [ ] Upfront about costs, limitations, shipping
  • [ ] Saves user steps wherever possible
  • [ ] Effort clearly put into quality
  • [ ] FAQs are real, current, and candid
  • [ ] Graceful error recovery
  • [ ] Apologizes when inconveniencing users

Step 10: Accessibility Quick Check (Chapter 12)

Krug's "low-hanging fruit" accessibility checks:

  • [ ] Text contrast: Sufficient contrast between text and background? No small, low-contrast type?
  • [ ] Form labels: Labels outside form fields (not floating inside)?
  • [ ] Text resizing: Does the page respond to browser text size changes?
  • [ ] Headings: Using semantic heading hierarchy (h1 → h2 → h3)?
  • [ ] Link distinction: Visited vs unvisited links visually different?
  • [ ] Keyboard navigation: Can all content be accessed via keyboard?

Output Format

Structure the audit report as:

# Krug UX Audit: [Site Name]
**URL:** [url]
**Date:** [date]
**Pages reviewed:** [list]

## Executive Summary
[2-3 sentence overall assessment]
[Overall grade: A/B/C/D/F with brief justification]

## Trunk Test Results
[Table with ✅/⚠️/❌ for each question]

## Krug's Laws Assessment
### Law 1: Don't Make Me Think
[Findings with specific examples]

### Law 2: Mindless Choices
[Findings with specific examples]

### Law 3: Omit Needless Words
[Findings with specific examples]

## Billboard Design
[Findings organized by principle]

## Navigation
[Findings organized by element]

## Home Page
[Four Questions assessment + element review]

## Mobile/Responsive
[If applicable]

## Courtesy & Goodwill
[Drains and builders identified]

## Accessibility
[Quick check results]

## Top 5 Issues (Prioritized)
1. [Most critical — what to fix first]
2. ...
3. ...
4. ...
5. ...

## Quick Wins
[Things that can be fixed in under an hour]

## Recommendations
[Specific, actionable changes ordered by impact]

Key Quotes to Ground the Audit

Use these Krug principles as anchoring references throughout:

  • *"If you can't make something self-evident, you at least need to make it self-explanatory."*
  • *"The most important thing you can do is understand the basic principle of eliminating question marks."*
  • *"If your audience is going to act like you're designing billboards, then design great billboards."*
  • *"Clarity trumps consistency."*
  • *"We don't read pages. We scan them."*
  • *"We don't make optimal choices. We satisfice."*
  • *"We don't figure out how things work. We muddle through."*
  • *"The reservoir of goodwill is not bottomless."*
  • *"People won't use your Web site if they can't find their way around it."*
  • *"FOCUS RUTHLESSLY ON FIXING THE MOST SERIOUS PROBLEMS FIRST."*

Important Notes

  • Be specific. Every finding must reference a concrete element on the page, not abstract theory.
  • Include screenshots/refs. When using browser snapshots, reference specific elements by their ref IDs.
  • Prioritize ruthlessly. Krug's mantra: fix the most serious problems first. Don't drown the report in minor issues.
  • No jargon without explanation. Krug wrote for everyone, not just UX professionals. Keep the audit readable.
  • Suggest, don't just criticize. Every problem should come with a practical fix suggestion.
  • Grade on real impact. A beautiful site with confusing navigation is worse than an ugly site that's easy to use.

平台分布

OpenClaw

86.05%
按下载量换算1,486

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills