Prompt
How do I review and approve intentional UI changes in visual tests?
Latest observation
Oct 27, 2026
OpenAI APIWeb search: off
To review and approve intentional UI changes in visual tests, use a workflow that makes it easy to compare the new screenshots against the baseline and then update the baseline only when the change is expected.
Typical workflow
-
Run the visual test suite
- Your test tool captures new screenshots and compares them to the stored baseline images.
-
Inspect the diffs
- Review side-by-side comparisons or overlay diffs.
- Check whether the change is expected, such as:
- a design update
- a text/content change
- a responsive layout adjustment
- a browser rendering difference
-
Approve the change if intentional
- If the visual difference is valid, mark it as approved in your visual testing tool.
- This usually updates the baseline image or stores the new screenshot as the approved version.
-
Reject or fix if unintentional
- If the difference is a regression, fix the code and rerun the tests.
Common ways to approve
- Via UI dashboard: most visual testing services let you open failed tests and click “Approve” or “Accept.”
- Via CLI or test runner: some tools let you accept new baselines from the command line.
- Via branch/PR workflow: approve changes in the visual review tool, then merge the PR once all checks pass.
Best practices
- Only approve changes after confirming they’re expected.
- Review diffs at the right screen sizes and browsers.
- Keep approvals tied to a pull request so the reason for the change is documented.
- Prefer reviewing with real examples rather than blindly updating all baselines.
Example with common tools
- Percy: review builds in the Percy dashboard and click Approve.
- Chromatic: inspect changes and click Accept changes.
- Applitools: mark differences as accepted in the test results view.
If you tell me which visual testing tool you’re using, I can give the exact approval steps for it.
Brands