How accessFlow code agent runs automatic and on-demand audits in GitHub

Note: accessFlow code agent is currently in beta and only available to eligible customers.

The accessFlow code agent AI Reviewer is built into your GitHub workflow to provide accessibility feedback as you work. It runs automatically in the background and can also be triggered manually when needed.

Automatic reviews

By default, the AI Reviewer works automatically the background.

Reviews are triggered:

Opening a pull request: The AI Reviewer starts scanning code changes as soon as a new pull request is created.
Updating code: Whenever you push new commits to an open PR, the AI Reviewer scans the latest changes.
Resolving issues: If a previously reported issue is fixed, the AI Reviewer updates its original comment with a green checkmark.

This helps you track progress without manually re-running scans.

On-demand reviews

You can manually trigger a review in GitHub at any time. This is useful when:

Automatic reviews are disabled in your .accesscode.yml file
You want to re-run a scan without pushing new code

To run a review on-demand

Navigate to your PR.
Go to the Reviewers section in the right-hand sidebar.
Select the settings icon (⚙️) next to Reviewers.
Select the accessFlow code agent app from the list.

The AI Reviewer will immediately run a new scan on your latest changes.

accessiBe Help Center

How accessFlow code agent runs automatic and on-demand audits in GitHub

Automatic reviews

On-demand reviews

See also

Was this article helpful?