Self-Healing Tests

A renamed class or restructured component should not take down a passing test. Self-Healing Tests catch a broken selector during execution and suggest a recovered locator so the test keeps running. This goes after the single biggest source of maintenance in UI suites: locators that break when the UI changes.

Who it’s for#

Teams running UI tests against an app that changes often, where selector churn creates a steady stream of false failures.

What it does#

• Locator recovery — When a selector breaks mid-run, AI suggests a recovered locator so the test continues.
• Healing record — See when healing happened across your runs instead of guessing.
• Suggestion review — Inspect the locator the AI proposed and decide whether to keep it.
• Maintenance reduction — Cut the rework caused by broken locators, the most common UI test failure.

How it works#

During execution, the platform watches for selectors that no longer match an element. When one breaks, AI proposes a replacement locator and the test keeps going rather than failing outright. Each outcome is recorded, so you can review where healing occurred and confirm the suggested locators.

Why it matters#

Broken locators are the leading reason UI suites need constant attention. Healing them in place keeps runs moving and gives you a record to review, so a changing UI no longer means a flood of failures to chase down.