Skip to main content
Bug scan finds bugs without any test cases. You give Peeps a URL, and an AI agent opens your app in a real browser, moves from page to page, and checks each one for problems. When it finishes, every bug it found is listed on the Bugs page with a severity, the page it was on, and evidence: a screenshot, the moment in the scan’s video recording, and the console or network data behind it.
The Bugs page listing nine bugs from the latest scan, grouped by type, with severity, page, evidence and status columns

What a scan checks

On every page it visits, the agent takes a snapshot of the page text, a screenshot, the browser console, and the network requests. It also runs an automated accessibility audit. From those it reports nine kinds of bug: The agent explores breadth-first and doesn’t revisit pages it has already checked. It reports real bugs only, not suggestions or style preferences.
The accessibility audit uses axe-core, which the agent loads into each page from cdnjs.cloudflare.com. If your app’s Content Security Policy blocks that domain, the scan can’t run the audit and won’t report accessibility bugs.

Severity

Each bug gets one of four severities:

How long a scan takes

Most scans finish in a few minutes. A scan stops after 15 minutes at most. As it nears the limit, the agent stops exploring and writes up what it has found so far, so you still get a report. Only one scan can run in a project at a time. To start another, wait for the current one to finish or cancel it.

Who can use it

Bug scan works on web projects. If you don’t see Bugs in the sidebar, Bug scan isn’t turned on for your organization yet. Get in touch to request access.

Next steps

Run a scan

Choose the URL, scope, and sign-in, then follow the scan live.

Review bugs

Read the evidence, triage each bug, and hand it to your coding agent.

Schedule scans

Scan automatically every day or week, and get notified when it’s done.