Complete AI Training
Sign inGet started

AI agent for web designers

Accessibility Audit Agent

Find every accessibility failure on key screens and confirm each fix

Quality controlComplianceIT and developmentProduct development
See how it worksGet access
Accessibility Audit Agent: what goes in, what the agent does and what you get

What it does

Accessibility problems pile up screen by screen: unlabeled buttons, low contrast, tiny touch targets, a focus order that jumps around. This agent walks through each screen of the app or site, using automated checks and a screen reader script. It tests labels, contrast ratios, touch target size, focus order, text scaling and form errors. It lists each failure with the element, the screen and the rule it breaks, and proposes a specific fix such as a label text or a color value. After the builder applies the fixes, it rechecks the same screens and a few related ones. It tracks the open count over time. The builder approves the changes. Edge case: a label is present but reads the same on five buttons, so the agent proposes unique labels.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
ApprovedYes, continueNo 1 STARTS WHEN Release candidate or monthly audit 2 USES A TOOL Open each key screen and run automated checks 3 USES A TOOL Walk the screen with a screen reader script andkeyboard path 4 DOES List failures with element, screen and rule 5 DOES Rank by how much the failure blocks a task 6 DOES Propose a specific fix for each failure 7 YOU APPROVE Builder approves the fixes to apply 8 USES A TOOL Builder applies the fixes; agent reruns the screensand related ones 9 CHECKS THE RESULT Did each fixed item pass and no new failures appear? If not: list the remaining and new failures with updatedfixes. Back to step 3. 10 RESULT Audit report with open count over time
Read the steps as a list
  1. Release candidate or monthly audit
  2. Open each key screen and run automated checks
  3. Walk the screen with a screen reader script and keyboard path
  4. List failures with element, screen and rule
  5. Rank by how much the failure blocks a task
  6. Propose a specific fix for each failure
  7. Builder approves the fixes to applyThe agent waits here for your OK.
  8. Builder applies the fixes; agent reruns the screens and related ones
  9. Did each fixed item pass and no new failures appear?If not: list the remaining and new failures with updated fixes. Back to step 3.
  10. Audit report with open count over time

How it decides

It measures each element against the chosen standard and ranks failures by how many users and tasks they block.

  • Require a 4.5 to 1 contrast ratio for normal text
  • Require touch targets of at least 44 points
  • Rank failures that block sign-in or checkout first
  • Check text at 200% size

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Standard (default WCAG 2.2 AA)
  • Key screens list
  • Contrast and touch size limits
  • Audit schedule
  • Issue tracker to write to

What keeps you in control

It always asks you first

  • Builder approves the changes
  • Design lead approves color changes

Hard limits

  • Never change designs, only propose changes
  • Never mark an item fixed without a recheck

It stops when

  • Done: all high-impact failures are closed
  • Stop: a design change is needed that the builder must decide

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensThe audit of 18 screens found 47 failures. Twelve blocked the checkout path, including a pay button with no label and a gray text at 2.8 to 1 contrast. The builder approved 40 fixes. The recheck passed 38 but the focus order on the address form jumped from the street to the state, and a new contrast failure appeared, so the check failed. Round two closed all.

More agents for web designers

Quality control · Creatives

Design System Drift Repair Agent

Working design files match the approved design library, without breaking layouts or intentional exceptions.

10 steps2 checks5 jobs Quality control · Creatives

Creative Handoff Packaging Agent

A complete delivery package that matches the client's checklist on the first try.

9 steps1 check5 jobs Quality control · Coding

Public Form Accessibility Agent

Every public form remains accessible after each revision.

7 steps2 checks5 jobs Planning · Decisions

Landing Page A/B Test Planning Agent

A single-change test plan that is launched correctly and read at the right time

11 steps2 checks6 jobs