What is the accessibility-testing skill?
The accessibility-testing skill is a SKILL.md in proffesor-for-testing/agentic-qe. It gives your AI coding agent a WCAG 2.2 quick reference: POUR principles, contrast ratios, a keyboard navigation test, axe-core automation for CI and a VoiceOver checklist. Its own text points to a11y-ally for full audits.
The accessibility-testing skill is a short quick reference card for WCAG 2.2 compliance work. It covers legal drivers such as ADA, Section 508 and the EU Directive, and it is meant for new features, release checks and audits.
It sets a five step default for your agent: apply the POUR principles, test with keyboard only, validate with screen readers, check color contrast, and automate with axe-core in the CI/CD pipeline.
Checks it puts in front of your agent
- All images have alt text, or alt="" for decorative ones.
- All form fields have labels.
- Color is never the only indicator.
- Focus is visible on every interactive element.
- Keyboard navigation works throughout the page.
Contrast and conformance levels
Normal text needs 4.5:1 at AA and 7:1 at AAA. Large text of 18pt and up needs 3:1 at AA and 4.5:1 at AAA, and UI components need 3:1. Level A is the minimum legal bar, AA the industry standard.
When to use a11y-ally instead
The skill states that for full WCAG auditing with axe-core, pa11y and Lighthouse, video accessibility and remediation, you should prefer the a11y-ally skill from the same repository. This one stays a quick card for basic patterns.
Who the accessibility-testing skill is for
Best for
- Teams that want accessibility checks inside an existing Playwright test suite.
- Release gates that need a fast WCAG 2.2 AA pass in CI/CD.
- Developers who want a keyboard and screen reader checklist next to the automated checks.
Not for
- Design reviews against WCAG 2.2 with severity ratings, with the accessibility-audit skill.
- Fixing the markup that a check flags, with the fixing-accessibility skill.
- Reviewing a Figma design or a written description against WCAG 2.1 AA, with the accessibility-review skill.
- Full WCAG audits with axe-core, pa11y and Lighthouse, where the a11y-ally skill from the same repository is preferred.
- Replacing manual testing: the skill says automation catches only 30 to 50% of issues.
How to install Accessibility Testing Skill: WCAG 2.2 Quick Reference for AI Agents
npm install -g agentic-qe
cd your-project
aqe init --autoA smaller Claude Code option is the plugin: /plugin install agentic-qe-fleet --marketplace proffesor-for-testing/agentic-qe.Check the requirements
The repository requires Node.js 22.13.0 or newer and npm 10.0.0 or newer, and works with the coding agent you already have configured.
Install and initialize
Run the three commands above.
aqe init --autoanalyzes your project and configures the default Claude Code surface, including MCP. Runaqe healthto check the setup.Pick another client if needed
Add a setup flag such as
--with-codex,--with-cursor,--with-copilot,--with-clineor--with-windsurftoaqe init --auto.Ask for an accessibility test
Request WCAG 2.2 compliance testing, a keyboard check or a screen reader checklist. The skill description covers ADA and Section 508 compliance, testing for disabilities and accessible application work.
Check that it loaded
Ask your agent to test a checkout page for WCAG 2.2 AA. Expect the five step flow, a keyboard test, contrast ratios checked against 4.5:1 and 3:1, and an axe-core run with the WCAG tags.
Key features
- Five step defaultApply POUR, test by keyboard, validate with screen readers, check contrast and automate with axe-core. The agent follows this order when asked to test accessibility.
- Quick a11y checklistAlt text, form labels, color as the only indicator, visible focus and keyboard navigation form a short checklist the agent can run on any page.
- WCAG 2.2 level tableA, AA and AAA are laid out with their targets: A as the minimum legal level, AA as the industry standard and AAA for specialized sites.
- POUR principles tableEach of the four POUR principles lists its key tests, from alt text, contrast and captions to keyboard access, clear labels, valid HTML and ARIA.
- Contrast thresholdsNormal text, large text and UI components each get an AA and AAA ratio, so the agent can check color pairs against a number.
- Keyboard navigation testsA Playwright test focuses every button, link, input, select and textarea, and a second test presses Tab and confirms the focus outline is not none.
- axe-core in CIAn AxeBuilder test with the wcag2a, wcag2aa and wcag22aa tags expects zero violations, and a checkout example fails the build on critical or serious impacts.
- Screen reader checklistA VoiceOver list covers the page title, heading order, landmarks, alt text, form labels, error messages and aria-live updates.
- Agent coordinationThe skill names qe-visual-tester, qe-test-generator and qe-quality-gate for running checks in parallel and blocking non-compliant builds.
Frequently asked questions
What is the accessibility-testing skill?
It is a SKILL.md in the proffesor-for-testing/agentic-qe repository that works as a quick reference card for WCAG 2.2 testing. It covers POUR principles, contrast ratios, keyboard checks, axe-core automation and a screen reader checklist for your AI coding agent, in a five step default order.
How do I install accessibility-testing?
Install the repository CLI with npm install -g agentic-qe, move into your project and run aqe init --auto. It needs Node.js 22.13.0 or newer. For clients other than Claude Code, add a flag such as --with-cursor or --with-codex to the init command.
What contrast ratios does accessibility-testing use?
Normal text needs 4.5:1 at AA and 7:1 at AAA. Large text of 18pt and up needs 3:1 at AA and 4.5:1 at AAA. UI components need 3:1. Your agent can compare each foreground and background pair against these numbers when it reviews a page.
Does accessibility-testing run axe-core?
Yes. It shows an AxeBuilder test from @axe-core/playwright using the wcag2a, wcag2aa and wcag22aa tags, expecting no violations. It keeps this path for existing Playwright suites or when you need Firefox and Safari coverage, and it prefers a11y-ally for new work.
How does it test keyboard access?
It uses two Playwright tests. The first focuses every button, link, input, select, textarea and tabindex element and confirms each one receives focus. The second presses Tab and checks that the focused element's computed outline is not none, so a visible focus indicator exists.
What is the difference between accessibility-testing and a11y-ally?
Accessibility-testing is a quick reference card with basic patterns. A11y-ally runs axe-core, pa11y and Lighthouse together, handles video accessibility and writes remediation guidance, and it drives the browser through the qe-browser engine. The accessibility-testing text itself says to prefer a11y-ally for full WCAG auditing.
Can automated tests prove a page is accessible?
No. The skill states that automated testing catches 30 to 50% of issues. It calls for manual testing with real assistive technology, such as VoiceOver, NVDA and JAWS, and for including users with disabilities in testing alongside the axe-core checks.
What license does accessibility-testing use?
It is MIT licensed. The proffesor-for-testing/agentic-qe repository uses the MIT license, and its GitHub description calls the project the Agentic QE Fleet, an open-source platform that is free to use, fork, build and contribute. The accessibility-testing skill lives at .claude/skills/accessibility-testing inside that repository.


