- Home
- Skills
- Frontend & Web Apps
- a11y-skills-toolkit
A11y Skills Toolkit
Frontend accessibility skills: audit real routes (WCAG 2.2 AA), fix in your stack, re-scan to prove it, add a Playwright + axe regression test.
Free
Works with the AI Agents you already use
a11y-skills-toolkit
Example session with this skill installed
Audit the login flow at /login and the dashboard at /dashboard. The app runs on port 3000. Give me a WCAG 2.2 AA report.
- Read your context and instructions
- Compiled the a11y--toolkit
- Generated the UI component
I have started the development server on port 3000 and scanned the requested routes using Playwright and axe-core. I also performed manual keyboard and focus checks on the login form.
Accessibility Audit Report
| Route | Criterion | Status | Severity | Finding |
|---|---|---|---|---|
/login | 1.1.1 Non-text Content | Fail | Serious | Login logo <img> lacks alt attribute. |
/login | 3.3.2 Labels or Instructions | Fail | Moderate | Password field lacks a visible <label>. |
/dashboard | 2.1.1 Keyboard | Pass | N/A | All sidebar nav items reachable via Tab. |
/dashboard | 2.4.7 Focus Visible | Fail | Critical | "Logout" button has outline: none with no replacement. |
/dashboard | 1.4.3 Contrast (Minimum) | Fail | Minor | Footer text ratio is 3.2:1 (Required 4.5:1). |
Summary
The audit found 1 Critical, 1 Serious, and 2 Moderate violations. The dashboard is currently unusable for keyboard-only users due to the missing focus indicators on primary actions.
Next steps
- Run
a11y-implementerto fix the missing focus rings in your Tailwind config. - Add
alttext to the branding assets incomponents/Header.tsx. - Re-scan the
/dashboardroute to verify the focus indicator fix.
Header.tsx
TSX · React component
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Accessibility compliance is often a bottleneck in the release cycle, usually discovered too late via manual audits. This toolkit automates WCAG 2.2 AA auditing and remediation directly within your development environment.
What it does
- Automated Scanning runs
axe-corevia Playwright across dynamic and static routes to catch low-hanging violations. - Manual Checklists provides guided keyboard, focus, and reflow verification steps for criteria that automation cannot detect.
- Remediation Engineering refactors existing React, Vue, Svelte, or Angular components using native HTML and ARIA patterns.
- Regression Testing generates Playwright test suites to ensure accessibility fixes persist in CI/CD pipelines.
- Reference Library offers instant access to WCAG 2.2 criteria, ARIA patterns, and severity models for EAA compliance.
How it works
- Audit a local or remote URL to generate a status report (Pass, Fail, Not Tested) mapped to specific success criteria.
- Implement fixes by providing the auditor's report to the implementer skill, which modifies your components.
- Verify every code change by triggering a re-scan of the affected route to prove the fix works.
- Protect the codebase by scaffolding an
a11y.spec.tsfile for your existing test runner.
Frameworks & tools
Works with React, Vue, Svelte, Angular, and Astro. Uses Playwright and axe-core for runtime analysis. Supports Tailwind CSS, CSS Modules, and major headless UI libraries like Radix and Headless UI.
Why this beats prompting it yourself
Generic LLMs often hallucinate ARIA attributes or suggest tabindex > 0, which makes accessibility worse. This skill uses a hardened reference library of APG-compliant patterns and enforces a "verify then claim" loop that requires actual scanner evidence before marking an issue resolved.
Use cases
- Auditing a production site for EAA (European Accessibility Act) readiness.
- Refactoring a legacy form to handle error identification and keyboard navigation correctly.
- Converting icon-only buttons into accessible interactive controls.
- Adding accessibility regression gates to a GitHub Actions pipeline.
Known limitations
Does not support native mobile apps (iOS/Android). Automated scans cannot detect logical focus order or meaningful alt text quality. Requires a running dev server for audits.
Web only; no native mobile support. Cannot verify 100% of WCAG manually. Requires Playwright/Node.js environment. Does not provide legal certification.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
5 uses
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 13 days ago
- Free with an Agensi account
Requires