- Home
- Skills
- Frontend & Web Apps
- Accessibility Audit Commander
Works with the AI Agents you already use
Accessibility Audit Commander
Structured WCAG 2.2 accessibility audits with severity-ranked findings and specific code fixes.
Free
Accessibility Audit Commander
Example session with this skill installed
- Read your context and instructions
- Compiled the accessibility audit commander
The audit of the React modal component is complete. This component has 1 Blocker and 2 Critical issues that prevent keyboard and screen reader users from successfully managing settings.
Accessibility Audit Report: Settings Modal
Scope: React Modal Component (inline snippet)
Summary: 1 Blocker, 2 Criticals, 0 Warnings. The component is currently inaccessible to keyboard-only and screen reader users due to a lack of focus management and missing labels.
Findings
| ID | Severity | WCAG Criterion | Location | What's wrong | Affected Users |
|---|---|---|---|---|---|
| A11Y-001 | Blocker | 2.4.3 Focus Order | div.modal | Focus is not trapped inside the modal when open. | Keyboard, Screen Reader |
| A11Y-002 | Critical | 3.3.2 Labels or Instructions | input | Input relies on placeholder text instead of a label. | Screen Reader, Cognitive |
| A11Y-003 | Critical | 4.1.2 Name, Role, Value | button | The close button has no accessible name, only a visual 'X'. | Screen Reader |
Remediation Code
Fix for A11Y-002 (Missing Label):
// Before
<input type="text" placeholder="Username" />
// After
<label htmlFor="username">Username</label>
<input id="username" type="text" placeholder="e.g. jdoe74" />
Fix for A11Y-003 (Accessible Name):
// Before
<button onClick={open}>X</button>
// After
<button onClick={close} aria-label="Close settings modal">X</button>
Remediation order
- Implement a focus trap using a library or
useEffectto prevent keyboard users from tabbing into the background (A11Y-001). - Add the
aria-labelto the close button so screen reader users know how to exit (A11Y-003). - Replace placeholders with proper
<label>elements for form persistence (A11Y-002).
Next steps
- Apply the focus trap logic to the modal container.
- Update the button and input JSX with the provided snippets.
- Verify the fix by tabbing through the open modal.
Connects securely to your tools. The creator never sees your data.
About this skill
Development teams often treat accessibility as a post-launch afterthought or a vague checklist. This skill runs a structured WCAG 2.2 audit directly on your frontend code, components, or live pages to find blockers before they reach production. It moves beyond generic advice by providing specific code-level fixes and ranking issues by their actual impact on users with assistive technology.
What it does
- Multi-pass auditing — Reviews interfaces through four distinct lenses: semantics, keyboard interaction, visual perception, and robustness.
- Severity ranking — Categorizes issues as Blocker, Critical, Warning, or Note based on how they impact real user tasks.
- Code-level remediation — Delivers exact code changes and ARIA implementations rather than general accessibility theory.
- Contrast verification — Uses a contrast checker to validate hex values against 4.5:1 and 3:1 WCAG thresholds.
- Assistive tech simulation — Evaluates how components behave for screen readers, keyboard-only users, and motion-sensitive users.
How it works
- Define scope — Identify the specific component, page, or code block that requires an audit.
- Execute audit passes — The agent runs a structured four-pass check starting with DOM semantics and ending with ARIA robustness.
- Analyze findings — Issues are mapped to specific WCAG 2.2 criteria and severity levels.
- Generate report — Receive a structured remediation plan including IDs, locations, and before/after code snippets.
Frameworks & tools
This skill works with any frontend stack including React, Vue, Svelte, and vanilla HTML/CSS. It utilizes scripts/contrast-checker.py for color analysis and references the WCAG 2.2 standard.
Why this beats prompting it yourself
Generic prompts often return vague "best practices" or hallucinate ARIA attributes. This skill follows a rigid four-pass framework and requires specific remediation code, ensuring you get actionable PR feedback instead of a generic checklist.
Use cases
- Pre-PR audits — Review a new UI component for keyboard traps and semantic errors before merging.
- Design system hardening — Audit core library components to ensure accessibility scales across the entire product.
- Compliance reporting — Generate a structured WCAG 2.2 status report for stakeholders or legal teams.
- Fixing broken UX — Identify why a specific form or modal is failing for screen reader users.
Known limitations
Audits are limited to the specific code or live scope provided. Script-based contrast checking requires raw hex values to be present in the source or provided manually.
Audits are scope-specific and do not cover entire sites unless all code is provided. Contrast checking requires explicit hex values. Logic ignores CSS-in-JS if values are generated at runtime.
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
This skill has been used 6 times.
Trust & safety
Security scan
Security scan passed 24 days ago
- Free with an Agensi account
Requires
Permissions
No special permissions declared or detected