What browser tests are
A browser test is one flow through your running app, recorded once and replayed whenever you ask.
What a browser test is
A browser test is a flow a person would follow: sign in, click a button, check that the right thing happened. You record it once. After that verifAIed replays it on demand, with a check after each step and a video of the whole run.
Unit coverage tells you which lines ran. A browser test tells you the screen still works.
Recorded tests and suite tests
- A recorded test is a flow recorded with verifAIed. It is saved as a plain
@playwright/testfile in your repository, and verifAIed can replay it. Every replay keeps a video, a trace and the code the browser reached. - A suite test mirrors one test in a Playwright suite you already have. Your own runner runs it and verifAIed reads the result. See Existing Playwright suites.
Three ways to make one
- Record it yourself. Start the studio, name the test and click through the flow. See Record a test.
- Let your coding agent record it. The agent drives your app itself and proves the flow passes. See Record with your agent.
- Track the Playwright suite you already have. See Existing Playwright suites.
The Browser Tests page
Every branch has a Browser Tests page. It is the page a branch opens on.
The tiles
| Tile | What it counts |
|---|---|
| Code reached | How many of your functions a real browser has run. |
| Tests | Every browser test on the branch. |
| Passing | Tests whose latest run on this branch passed. |
| Failing | Tests whose latest run on this branch failed. |
| Not run | Tests this branch has never run, plus drafts. |
| Recorder | On when a studio is running on your machine, Off otherwise. |
| Stale links | Tests that contain a linked copy of another test that has been re-recorded since. |
The feature tree
Tests are grouped into folders, one per feature, nested as deep as you like. Each folder shows a summary of the tests inside it and has its own Run button. Click a test to open its page, with its steps, its runs and its video.
On a feature branch, the New & edited / All tests toggle switches between the tests this branch added or changed and every test. A test the branch added carries a new badge, and one it changed carries a modified badge.
What a row tells you
| Status | Meaning |
|---|---|
| Pass | The latest run on this branch passed. |
| Fail | The latest run on this branch failed. |
| Never run | The test is recorded, but this branch has not run it yet. |
| Draft | The test has a name but no recording yet. |
| Queued | A run or a recording is waiting for the studio. |
| Running | The studio is running or recording it right now. |
A row can also carry one of these marks:
- Recording live: somebody is recording this flow right now. Click it to open the recording.
suite: the test is tracked from your own Playwright suite.- not seen in latest run: the newest run of your suite did not contain this test.
- A note that a linked copy of another test is out of date. The next run updates it. See Inserted tests.
On a test's own page, a run that never reported a result is listed as Abandoned.
Below the tree
- Browser coverage: which functions a real browser ran. See Code reached.
- Visual Verdicts: the results of visual checks on this branch. See Visual checks.
- Other activity: every other browser session on the branch, such as visual checks, page audits, manual recording sessions and runs of your own suite.
Browser tests and done
A recorded test that this branch has never run, or whose latest run failed, keeps the branch from being done. See Definition of done.