Accessibility Audit logo

Accessibility Audit

CommunityPopular
mohitagw15856
accessibility-audit

Generate a WCAG 2.2 accessibility audit checklist and remediation suggestions for any UI or design. Use when asked to audit for accessibility, check WCAG compliance, review a design for a11y issues, or create an accessibility remediation plan. Produces a prioritised checklist with pass/fail assessments and specific fixes.

Overview

Publishermohitagw15856
Repositorypm-claude-skills
Skill nameaccessibility-audit
Stars
1.4K
Forks
240
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by mohitagw15856 on GitHub. Read the source before you install it.

Installation

Install the Accessibility Audit AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/mohitagw15856/pm-claude-skills.git /tmp/pm-claude-skills
mkdir -p .claude/skills
cp -r /tmp/pm-claude-skills/exports/openclaw/accessibility-audit .claude/skills/accessibility-audit
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Accessibility Audit in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Accessibility Audit on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Accessibility Audit is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Accessibility Audit Skill

This skill produces a structured accessibility audit based on WCAG 2.2 guidelines. It covers visual, motor, cognitive, and screen reader accessibility — with prioritised remediation for each issue found.

Required Inputs

Ask the user for these if not provided:

  • What is being audited (screen, component, full product, design spec)
  • Description or image of the UI
  • Target WCAG level (A / AA / AAA — default to AA, which is the legal standard in most jurisdictions)
  • Known assistive technology users? (Yes/No — if yes, which: screen reader / switch access / voice control / magnification)
  • Platform (Web / iOS / Android / Desktop app)

Programmatic Helper

Contrast ratios cannot be eyeballed. The AA line sits at 4.5:1, and #777777 on white is 4.478 (fails) while #767676 is 4.54 (passes) — no amount of looking at a screenshot separates those. Compute them:

bash
npx --yes notugly fix "#8ab4f8" "#ffffff"     # ratio, APCA, and the nearest passing colour
npx --yes notugly onepager <url> --out review.html   # every pairing, printable
npx --yes notugly vision                      # which colours merge for colour-blind viewers

notugly fix returns the ratio, the APCA lightness contrast, and the closest colour to the one already chosen that passes — same hue, same chroma. Paste those numbers into the tables below rather than estimating them.

Deterministic, zero dependencies, and no model call — so it costs nothing to run and gives the same answer every time.

For 1.4.3 Contrast (Minimum) and 1.4.11 Non-text Contrast, every row in the remediation table should carry a measured ratio, not an assessment. If the user supplied a screenshot rather than hex values, say so explicitly in the audit — an inferred ratio is not an audit finding.

Output Structure


Accessibility Audit: [Component or Screen Name]

Target standard: WCAG 2.2 Level [AA] Platform: [Platform] Date: [Date]


Audit Summary

CategoryIssues FoundCriticalModerateMinor
Perceivable
Operable
Understandable
Robust
Total

Overall compliance status: ✅ Compliant / 🟡 Minor issues / 🔴 Fails AA standard


Perceivable

1.1 Text Alternatives

  • All images have descriptive alt text (not filename or "image")
  • Decorative images have alt="" to be skipped by screen readers
  • Icons without visible labels have accessible names
  • Complex images (charts, diagrams) have extended descriptions

Issues found: [List specific issues or "None"]

1.3 Adaptable

  • Content structure uses semantic HTML (headings, lists, landmarks) — not just visual formatting
  • Reading order in DOM matches visual order
  • Form inputs have associated labels (not placeholder text as label)
  • Data tables have proper headers and scope

Issues found:

1.4 Distinguishable

  • Text contrast ratio ≥ 4.5:1 (normal text) or ≥ 3:1 (large text 18px+)
  • UI component contrast ratio ≥ 3:1 against background
  • Information is not conveyed by colour alone
  • Text can be resized to 200% without loss of content
  • No content that auto-plays audio

Issues found:


Operable

2.1 Keyboard Accessible

  • All interactive elements are reachable by keyboard (Tab key)
  • No keyboard traps
  • Custom components have keyboard interactions (arrow keys for menus, Escape to close modals)
  • Skip navigation link available for pages with repeated navigation

Issues found:

2.4 Navigable

  • Focus is visible at all times (not removed with outline: none without replacement)
  • Focus order is logical and predictable
  • Page/screen has a descriptive title
  • Link text is descriptive (not "click here" or "read more")
  • Headings are hierarchical (H1 → H2 → H3, no skips)

Issues found:

2.5 Input Modalities

  • Touch targets are at least 44x44px
  • No functionality requires complex gestures (pinch, multi-touch) without a simple alternative
  • Motion or dragging interactions have button alternatives

Issues found:


Understandable

3.1 Readable

  • Language of the page is set (lang attribute)
  • Unusual words, abbreviations, or jargon are explained

3.2 Predictable

  • Navigation is consistent across screens
  • Components behave consistently (same button does the same thing)
  • No unexpected context changes on focus or input

3.3 Input Assistance

  • Error messages identify the field and describe the error in plain language (not just "Invalid input")
  • Required fields are labelled (not just with colour or asterisk alone)
  • Forms provide suggestions for correcting errors where possible

Issues found:


Robust

4.1 Compatible

  • HTML is valid and well-structured
  • ARIA roles and attributes are used correctly (not to fix broken semantics)
  • Status messages (success, error, loading) are announced to screen readers without focus change

Issues found:


Prioritised Remediation List

PriorityIssueWCAG CriterionFixEffort
🔴 Critical[Issue][e.g. 1.4.3 Contrast][Specific fix][Low/Med/High]
🟡 Moderate[Issue]
🟢 Minor[Issue]

Priority definitions:

  • 🔴 Critical: Blocks access for users with disabilities. Legal risk. Fix before launch.
  • 🟡 Moderate: Significant friction. Fix in next sprint.
  • 🟢 Minor: Best practice. Address in roadmap.

Quick Wins (Fix in < 1 hour)

[List any issues that are trivially fixable — e.g. adding alt text, fixing contrast with a colour swap, adding a lang attribute. These are easy to ship immediately.]


Testing Recommendations

  • Manual keyboard test: Tab through the entire flow. Can you complete every task without a mouse?
  • Screen reader test: VoiceOver (Mac/iOS), NVDA or JAWS (Windows). Is every piece of content and every action accessible?
  • Colour contrast check: Use Stark (Figma plugin) or WebAIM Contrast Checker
  • Automated scan: Axe DevTools or Lighthouse accessibility audit (catches ~30% of issues automatically)

Quality Checks

  • Issues are mapped to specific WCAG criteria
  • Every critical issue has a specific fix recommendation
  • Quick wins are separated from larger fixes
  • Effort estimates are included for prioritisation
  • Testing recommendations are included

Anti-Patterns

  • Do not rely solely on automated scanning tools — automated checks catch ~30% of issues; manual keyboard and screen reader testing is required
  • Do not label an issue "minor" simply because it only affects a small percentage of users — for those users it may block all access
  • Do not add ARIA roles to fix broken semantics — use correct semantic HTML first; ARIA is a last resort
  • Do not confuse colour contrast of text with colour contrast of UI components — they have different minimum ratios (4.5:1 vs 3:1)
  • Do not audit only the happy path — error states, empty states, and loading states must also meet accessibility requirements

Example Trigger Phrases

  • "Audit this design for accessibility"
  • "Check WCAG compliance for [screen/component]"
  • "Give me an a11y audit of [UI description]"
  • "What accessibility issues does this design have?"

Frequently asked questions

What does the Accessibility Audit AI skill do?

Generate a WCAG 2.2 accessibility audit checklist and remediation suggestions for any UI or design. Use when asked to audit for accessibility, check WCAG compliance, review a design for a11y issues, or create an accessibility remediation plan. Produces a prioritised checklist with pass/fail assessments and specific fixes.

Why use Accessibility Audit on TypingMind?

Because you install it once and use it with any model. Accessibility Audit is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Accessibility Audit in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/mohitagw15856/pm-claude-skills/tree/main/exports/openclaw/accessibility-audit. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Accessibility Audit?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Accessibility Audit?

As many as you like. As long as a model supports skills, you can use Accessibility Audit with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Accessibility Audit AI skill free?

Yes. It is published on GitHub by mohitagw15856 under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇