What is the accessibility-audit skill?
The accessibility-audit skill is a short instruction file in the Owl-Listener designer-skills pack, inside its design-systems plugin. It makes your agent audit a design or build against WCAG 2.2 and write findings with a severity from Critical to Enhancement, plus criterion, impact, remediation steps and code examples.
The accessibility-audit skill is one of eleven skills in the design-systems plugin of Owl-Listener's designer-skills pack. Its SKILL.md casts the agent as an expert in digital accessibility, WCAG guidelines and inclusive design, and asks for audits that name barriers and explain how to remove them.
Its scope is an interface that exists today, whether a design or a build. Planning future sessions with assistive technology users is a different job, and the skill description sends you to accessibility-test-plan in the prototyping-testing plugin for that.
What an audit finding contains
Each finding follows one fixed format, so the list can move straight into a ticket tracker. The skill asks for a description, location, WCAG criterion, severity, impact, remediation steps and code examples. Severity runs on four levels.
- Critical: blocks access entirely.
- Major: significant difficulty.
- Minor: an inconvenience with workarounds.
- Enhancement: an improvement beyond compliance.
The four WCAG principles it audits against
- Perceivable: text alternatives, captions, adaptable content and color contrast.
- Operable: keyboard access, time limits, no seizures, navigation and input modalities.
- Understandable: readable, predictable content with input assistance.
- The fourth principle: assistive technology compatibility, semantic markup and ARIA.
Who the accessibility-audit skill is for
Best for
- Assessing a finished design or a live build before launch and getting a prioritized list.
- Designers and developers who need findings tied to a named WCAG criterion.
- Teams that want the same finding format on every audit.
Not for
- Planning sessions with assistive technology users. Use accessibility-test-plan instead.
- Editing your code. The skill defines findings and remediation steps. Related: fixing-accessibility.
- Replacing real testing. The skill itself says to test with real assistive technologies.
Use it with UiChemy
Run the audit on the Figma design or the built page before it ships. UiChemy Figma to WordPress converts a design into a WordPress site, so contrast and focus problems caught in the design are cheaper to fix. Click Get Started on uichemy.com. Related: web-accessibility-skill.
How to install Accessibility Audit Skill: WCAG 2.2 Findings by Severity
/plugin marketplace add Owl-Listener/designer-skillsRegister the marketplace
Type the command above into your agent's chat. It only tells the agent where the skills live, and nothing installs yet. The README writes this step for Claude Code.
Install design-systems
Run
/plugin install design-systems@designer-skills, or type /plugin, open the Discover tab, tick the plugin and press enter. The accessibility-audit skill ships inside it.Using Gemini CLI instead
The README clones the repository to a temporary folder and copies its .gemini/extensions contents into .gemini/extensions in your project root.
Request an audit
Name the screen or build and ask for an accessibility audit against WCAG. Say whether you are sharing a design or code.
Check that it loaded
Ask for an audit of one screen. The reply should sort findings into the four severity levels and give location, criterion, impact and remediation for each issue.
Key features
- WCAG 2.2 framingThe skill organizes the audit around the four POUR principles of WCAG 2.2, starting with Perceivable, Operable and Understandable, and ending with assistive technology compatibility.
- Perceivable checksLooks at text alternatives, captions, adaptable content and color contrast on the interface under review.
- Operable checksCovers keyboard access, time limits, seizure risk, navigation and the input modalities a person might use.
- Understandable and compatibility checksChecks that content is readable and predictable, that inputs get assistance, and that markup and ARIA work with assistive technology.
- Four severity levelsEvery issue is rated Critical, Major, Minor or Enhancement, so the worst barriers surface first.
- Fixed issue formatEach finding lists description, location, WCAG criterion, severity, impact, remediation steps and code examples.
- Real assistive technology testingBest practices include testing with real assistive technologies and involving users with disabilities when possible.
- Devices, browsers and statesThe audit runs across devices and browsers and looks at static states as well as interactive ones.
- Impact-based priorityFindings are prioritized by severity and by user impact.
Frequently asked questions
What is the accessibility-audit skill?
It is a skill in Owl-Listener's designer-skills pack, inside the design-systems plugin. It tells an AI agent to audit an existing design or build against WCAG 2.2 and report barriers with severity ratings, criteria, impact and remediation steps. It assesses what exists now rather than planning future testing.
How do I install accessibility-audit?
Add the marketplace with /plugin marketplace add Owl-Listener/designer-skills, then install the design-systems plugin with /plugin install design-systems@designer-skills. You can also tick it on the Discover tab of /plugin. Gemini CLI users follow the README's clone and copy steps.
Which WCAG version does accessibility-audit use?
The SKILL.md names WCAG 2.2 and its four principles, known as POUR, with the fourth one covering assistive technology compatibility. It does not list individual success criteria. Each finding asks for the relevant WCAG criterion, so the agent supplies it for every issue it reports.
What severity levels does the skill use?
Four levels. Critical means access is blocked entirely, Major means significant difficulty, Minor means an inconvenience with workarounds, and Enhancement means an improvement beyond compliance. The skill also says to prioritize findings by severity and by user impact when you decide what to fix first.
Which AI coding agents can use accessibility-audit?
The pack's README is written for Claude Code and Gemini CLI. Claude Code installs through the plugin marketplace, and each plugin is also a Gemini CLI extension. The README does not name other agents, so treat anything else as untested.
How is accessibility-audit different from accessibility-test-plan?
Accessibility-audit assesses a design or build that exists now and produces findings. Its description says to use accessibility-test-plan, in the prototyping-testing plugin, for planning future sessions with assistive technology users. One reports barriers today, the other prepares research for later.
Does the accessibility-audit skill fix the problems it finds?
The SKILL.md asks for remediation steps and code examples inside each finding, not for edits to your files. After the report you can ask the agent to apply the fixes. Related: fixing-accessibility.
What license is the designer-skills repository under?
The repository is MIT licensed according to GitHub. That covers the accessibility-audit skill folder and the rest of the pack. Check the repository before you redistribute the folder inside a product, Related: impeccable-audit.