AI auto-fix (beta)

​

Test maintenance is time consuming

End-to-end tests break often. Maintenance of a test suite can be more time consuming than the set-up and scripting combined. It is one of the reasons we started Octomind.

We gave our users access to the first version (beta) of our AI auto-fix feature.

This feature is in beta mode. We want you to use it and give us feedback on anything you liked or missed, let us know how successful you were.

This is how to proceed when you diagnosed a broken test.

1. Go to test report , into test result detail and click on the auto-fix tab.
2. Have the Agent attempt the auto-fix by clicking on the attempt auto-fix button.

3. Agent will take some time to explore the fix. You will see its progress in the AI Agent status bar.

4. Click on see proposed solution button.

5. Review proposed fix step by step (red steps are original - broken steps, green are new - proposed fixes). Approve or decline the auto-fix suggestion.

6. When you approved the fix, the new steps of the test case will be saved. If you didn’t like the suggestions, you can:

• have the Agent retry the auto-fix
• decline the auto-fix and edit the test case manually
• try editing the prompt