AI code remediation

Help developers fix accessibility issues early in the development process with AI code remediation. It provides actionable suggestions directly in pull requests (PR), improving both compliance and user experience without disrupting workflow.

On this page:

• Overview
• Commit-time code review with AI
• Supported violations
• Known limitations

Overview

Our solution uses AI to suggest code-level fixes when HTML accessibility issues appear during the commit process. Integrated with GitHub, it scans committed code using the CA11y tool. If it detects a violation, it posts a PR comment with a clear description, code example, and relevant documentation.

This approach supports faster resolution of issues and reduces reliance on post-development audits.

Commit-time code review with AI

To analyze committed HTML code and suggest fixes before merge:

1. Commit your HTML code to GitHub.
2. The CA11y tool scans the commit for accessibility violations.
3. If a supported issue exists, AI Code Remediation analyzes the code.
4. The system posts a comment in the PR with:
   • A description of the issue.
   • A recommended code fix.
   • The line of code and suggested fix which can be committed directly into the working file 

Known limitations

Current limitations for AI code remediations include:

• HTML support only
• Works with GitHub only
• Does not support AI conversations or full audits in PRs.