Read a test run report

Every run produces a report — the step-by-step record of what happened, with screenshots and timings. This guide is about reading one and finding old ones. To share a report, see How to share a test run report.

Find a run

Open Reports from the top nav. The list has a toggle between two histories:

Tests — single-test runs.
Test Sets — batch runs (a test set or tag run).

Narrow the list with the pills, the search box, a date-range picker, and — on the Tests view — a tag filter. The list is paginated; older runs are a page back, not gone (until they age out — see retention).

The two views filter on different things, because they record different things:

Tests — All / Passed / Failed / Running. A single run is Passed only when at least one step ran and none failed. A run in which no step ran at all is Failed, and its report says Nothing ran — zero failures is not the same as a pass.
Test Sets — All / Passed / Passed, with skips / Failed / Nothing ran / Stopped. These are the run’s verdict, worked out against the tests you asked for rather than the answers that came back:

Verdict	What it means
Passed	Every test asked for ran and passed.
Passed, with skips	Everything that ran passed; the only tests that didn’t run were ones you deliberately marked to skip.
Failed	Something failed — or a test asked for never reported, or had no steps to run. That includes a run that ended without finishing: a local run whose agent did not come back within 90 seconds (its status reads `disconnected`), a run our server lost, and a run that waited to start past its limit (2 hours in our queue, 24 hours for an agent). If the run had begun, each test that never answered fails by name and says why. Either way the report shows one line under the verdict saying why the run ended — for a run that never started, the only reason there is — and the `origin` says whose side it was on.
Nothing ran	No test was tested: an empty set, or every test was one you marked to skip. (A test skipped because it has no steps makes the run Failed instead — nobody asked for that skip.) Never shown as a pass.
Stopped	Someone stopped the run. Not a judgement about the tests.

Read a single-test report

The report header shows the test name (or Quick Run if it wasn’t a saved test), an overall pass/fail icon, the duration, and when it ran. Below that is the step list. Each step shows:

A status icon — passed, failed, or skipped.
The step description and how long it took.
A screenshot toggle, plus a button to open it full-size in the lightbox.
On a failure: the error message (expandable, with a Copy button) and a diagnosis callout that explains why the locator or assertion didn’t land.

What a step’s screenshot shows

A screenshot is taken after every step. When the step acted on or checked an element — a click, typing into a field, a select, a check against an element, a wait for one — that element is scrolled to the middle of the screen first and framed in the screenshot, so on a phone-sized run it is in the picture even when it sat below the fold. A text check that names no element frames the element the text was found in. Steps with no element — opening a page, scrolling, a URL or title check — show the page as it was. Visual steps keep their own screenshots and baselines.

Click any screenshot to open the lightbox; the step’s element is framed there too. Arrow keys (←/→) move between steps, Esc closes it, and you can download the full-size image to drop into a ticket.

Read a batch (test-set) report

A test-set or tag run rolls up into a batch report headed by its verdict badge — the same five verdicts as above — then stat cards — Passed / Failed / Skipped / Total — across every test, plus the browsers it ran on and the execution target (Cloud or Local). Below the rollup is the per-test breakdown.

The badge and the counts can look like they disagree, and that is on purpose: a run of one test that was skipped for having no steps shows Failed, with Passed 0 / Failed 0 / Skipped 1. Nothing failed, but nothing was tested either, and the verdict is computed against what you asked for.

Each row shows the test, its status badge, the per-test result, duration, and any error. Click a row to drill into that single test’s full report — same step list and screenshots as above.

From a batch report you can also Rerun the whole set or Rerun Failed — see Run tests and rerun the failures.

Runs that started signed in

A run that injected a session wears a Session: stored or Session: per run badge in its header, and each test in it records which it used. When the session no longer worked, the first step on the session’s host fails with an “auth expired — refresh the session cookie” message rather than a locate failure further down.

How long reports stick around

Run history is kept per plan tier:

Plan	Retention
Free	14 days
Starter	90 days
Pro	180 days
Team	2 years
Enterprise	Unlimited

Old runs are pruned automatically by a nightly cleanup, and downgrading shrinks your window (the downgrade preview shows what you’d lose). The full story on what’s stored and for how long is in Data handling.

Share a test run report — links, emails, and exports
Run tests and rerun the failures — rerun from a report
Capture and update visual baselines — reviewing visual diffs in a report
Test statuses — the Draft / Ready / Archived badge on each test