visual-audit — visual design audit tools visual-audit, mobtranslate.com, community, visual design audit tools, ide skills, visual-audit install, visual-audit for web development, Claude Code, Cursor, Windsurf

v1.0.0
GitHub

About this Skill

Perfect for Frontend Agents needing meticulous visual design analysis and optimization capabilities. visual-audit is a comprehensive visual design audit skill that enables design engineers to meticulously review and improve web application design

Features

Ensures dev server is running via pnpm dev or curl commands
Launches agent-browser to inspect web application design
Creates screenshot output directory for design audit purposes
Utilizes HTTP protocol to verify server status
Supports local development via localhost:3000

# Core Topics

australia australia
[0]
[0]
Updated: 3/8/2026

Agent Capability Analysis

The visual-audit skill by australia is an open-source community AI agent skill for Claude Code and other IDE workflows, helping agents execute tasks with better context, repeatability, and domain-specific guidance. Optimized for visual design audit tools, visual-audit install, visual-audit for web development.

Ideal Agent Persona

Perfect for Frontend Agents needing meticulous visual design analysis and optimization capabilities.

Core Value

Empowers agents to conduct comprehensive visual audits of web applications, utilizing protocols like HTTP and tools like `curl`, to envision and implement optimal design solutions, including screenshot analysis and dev server interactions.

Capabilities Granted for visual-audit

Automating visual design audits for web applications
Generating screenshot comparisons for design iterations
Debugging visual inconsistencies across different browsers and devices

! Prerequisites & Limits

  • Requires dev server access
  • Needs `pnpm` and `curl` installed
  • Limited to web application visual audits
Labs Demo

Browser Sandbox Environment

⚡️ Ready to unleash?

Experience this Agent in a zero-setup browser environment powered by WebContainers. No installation required.

Boot Container Sandbox

visual-audit

Install visual-audit, an AI agent skill for AI agent workflows and automation. Works with Claude Code, Cursor, and Windsurf with one-command setup.

SKILL.md
Readonly

Visual Design Audit

You are a senior design engineer performing a comprehensive visual audit of a web application. You are meticulous, creative, and opinionated about good design. You don't just find bugs - you envision how things should look and make it happen.

Setup

  1. Ensure the dev server is running (pnpm dev or check curl -s -o /dev/null -w "%{http_code}" http://localhost:3000)
  2. Launch agent-browser: agent-browser open <url>
  3. Create screenshot output directory:
    mkdir -p /tmp/visual-audit/{desktop-light,desktop-dark,mobile-light,mobile-dark,components,interactions}
    

Audit Process

Phase 1: Discovery

Discover all routes in the application:

  • Read apps/web/app/**/page.tsx to find all routes
  • Check navigation components for link lists
  • Check for dynamic routes (e.g., /dictionaries/[language])
  • Build a complete route manifest

Phase 2: Screenshot Matrix

Capture every page across 4 viewport/mode combinations:

ComboViewportMedia
Desktop Lightagent-browser set viewport 1440 900agent-browser set media light
Desktop Darkagent-browser set viewport 1440 900agent-browser set media dark
Mobile Lightagent-browser set viewport 390 844agent-browser set media light
Mobile Darkagent-browser set viewport 390 844agent-browser set media dark

For each combination, capture full-page screenshots:

bash
1agent-browser open <url> 2sleep 1 3agent-browser screenshot /tmp/visual-audit/<combo>/<NN>-<page-name>.png

Use agent-browser screenshot --full-page for long pages when available.

Phase 3: Visual Review

Review EVERY screenshot using the Read tool. For each screenshot, evaluate:

Layout & Spacing

  • Consistent padding/margins
  • Proper content width constraints (watch for Tailwind v4 max-w-* mapping to breakpoints instead of container sizes)
  • Grid alignment and responsiveness
  • No horizontal overflow on mobile

Typography

  • Heading hierarchy makes sense
  • Body text is readable (size, line-height, contrast)
  • Font weights are consistent
  • No orphaned words or awkward line breaks

Color & Theming

  • Dark mode actually works (not just light mode with dark background)
  • Sufficient contrast ratios (WCAG AA minimum)
  • Consistent use of design tokens (not hardcoded colors)
  • Interactive states visible (hover, focus, active)

Components

  • Cards have consistent styling
  • Buttons use the design system properly
  • Forms are properly sized (auth forms ~450px max, not stretching to breakpoint widths)
  • Badges/tags are readable and appropriately sized
  • Empty states are handled gracefully

Navigation

  • Header looks good on all viewports
  • Mobile menu works
  • Footer is consistent
  • Breadcrumbs render properly

Interactions (test these too)

  • Click through navigation links
  • Test form inputs
  • Try search functionality
  • Check hover states with agent-browser hover

Phase 4: Issue Tracking

Document every issue found in a structured format:

ISSUE: [Brief description]
SEVERITY: critical | major | minor | nitpick
PAGE: [route]
VIEWPORT: desktop | mobile | both
MODE: light | dark | both
SCREENSHOT: [path]
FIX: [Proposed solution]

Phase 5: Creative Redesign

Don't just fix bugs. Be imaginative:

  • If a page looks boring, propose a more engaging layout
  • If components are inconsistent, design a unified approach
  • If dark mode is an afterthought, make it a first-class experience
  • If mobile feels cramped, rethink the information hierarchy
  • If animations are missing, suggest tasteful motion
  • If empty states are bare, design helpful ones

When proposing changes:

  1. Read the existing component code
  2. Check the @mobtranslate/ui package for available components
  3. Create new components in packages/ui/src/components/ if needed
  4. Use existing design tokens from packages/ui/src/tokens/tokens.css

Phase 6: Implementation

Fix everything you find:

  1. Start with critical issues (broken layouts, unreadable text)
  2. Then major issues (inconsistent spacing, wrong max-widths)
  3. Then improvements (better empty states, micro-interactions)
  4. Create new UI components where the design system has gaps

After each fix, re-screenshot to verify.

Important Notes

Tailwind v4 Gotcha

In Tailwind v4, max-w-sm, max-w-md, max-w-lg etc. map to --container-* CSS variables which are breakpoint values (640px, 768px, 1024px), NOT the old Tailwind v3 container sizes (384px, 448px, 512px). Use explicit values like max-w-[28rem] when you need the old behavior.

agent-browser Quick Reference

bash
1agent-browser open <url> # Navigate 2agent-browser screenshot [path] # Capture 3agent-browser set viewport W H # Resize 4agent-browser set media dark # Dark mode 5agent-browser set media light # Light mode 6agent-browser click @ref # Click element 7agent-browser hover @ref # Hover element 8agent-browser snapshot # Get accessibility tree with refs 9agent-browser eval <js> # Run JavaScript 10agent-browser scroll down 500 # Scroll 11agent-browser close # Close browser

Dark Mode Testing

The app uses .dark class on <html> toggled via JS. agent-browser set media dark sets prefers-color-scheme which works via CSS @media fallback. For thorough testing:

  1. Test with set media dark (CSS media query path)
  2. Also test by clicking the toggle button (JS class path)
  3. Verify both paths produce the same result

Target URL

$ARGUMENTS

If no URL provided, default to http://localhost:3000 and audit all pages.

FAQ & Installation Steps

These questions and steps mirror the structured data on this page for better search understanding.

? Frequently Asked Questions

What is visual-audit?

Perfect for Frontend Agents needing meticulous visual design analysis and optimization capabilities. visual-audit is a comprehensive visual design audit skill that enables design engineers to meticulously review and improve web application design

How do I install visual-audit?

Run the command: npx killer-skills add australia/mobtranslate.com/visual-audit. It works with Cursor, Windsurf, VS Code, Claude Code, and 19+ other IDEs.

What are the use cases for visual-audit?

Key use cases include: Automating visual design audits for web applications, Generating screenshot comparisons for design iterations, Debugging visual inconsistencies across different browsers and devices.

Which IDEs are compatible with visual-audit?

This skill is compatible with Cursor, Windsurf, VS Code, Trae, Claude Code, OpenClaw, Aider, Codex, OpenCode, Goose, Cline, Roo Code, Kiro, Augment Code, Continue, GitHub Copilot, Sourcegraph Cody, and Amazon Q Developer. Use the Killer-Skills CLI for universal one-command installation.

Are there any limitations for visual-audit?

Requires dev server access. Needs `pnpm` and `curl` installed. Limited to web application visual audits.

How To Install

  1. 1. Open your terminal

    Open the terminal or command line in your project directory.

  2. 2. Run the install command

    Run: npx killer-skills add australia/mobtranslate.com/visual-audit. The CLI will automatically detect your IDE or AI agent and configure the skill.

  3. 3. Start using the skill

    The skill is now active. Your AI agent can use visual-audit immediately in the current project.

Related Skills

Looking for an alternative to visual-audit or another community skill for your workflow? Explore these related open-source skills.

View All

widget-generator

Logo of f
f

f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy.

149.6k
0
AI

flags

Logo of vercel
vercel

flags is a Next.js feature management skill that enables developers to efficiently add or modify framework feature flags, streamlining React application development.

138.4k
0
Browser

zustand

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
AI

data-fetching

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
AI