Understanding Run Results
This page covers the run detail page (/runs/{id}): everything you see
while a test is queued, running, and after it finishes, including exactly
what each status means and how the live-updating page works.
Starting a draft run
A run you've just created from the New Playtest form sits in draft status. Nothing has executed and no credit has been spent yet. The run detail page for a draft shows a Run this test button along with a note on exactly how many credits it will cost (1 for a normal run, 3 if Chaos Mode is enabled).
Clicking Run this test:
- Deducts the credit cost from your balance immediately (or shows an out of credits message with a link to buy more, if your balance is too low; see Billing & Credits).
- Moves the run to queued status and hands it off to VeriWasp's background job queue for execution.
The five statuses
| Status | Meaning |
|---|---|
draft | Created, steps saved, nothing executed, no credit spent. |
queued | Credit deducted, waiting for a worker to pick it up. |
running | A worker has started executing the steps, in order. |
completed | Every step finished; the overall run is considered a pass. |
failed | Either the browser/execution itself hit a fatal error, or at least one step failed. |
A run is marked failed rather than completed if any single step
reports a failure. There's no partial-credit state. If you want to know
specifically which step(s) failed and why, that detail lives in each
step's own result and in the run's issue list (see
Issues & Accessibility Scanning), not just in
the overall status badge.
Live updates while queued or running
While a run is in queued or running status, the page automatically
polls the server every 2 seconds and refreshes in place. You don't need
to manually reload the page to see progress. What updates live:
- The status badge itself, as it moves from queued to running to its final state.
- Each step's individual result (passed/failed, with its screenshot and any error message), appearing as soon as that step finishes. You can watch a long scenario complete step by step in near real time rather than waiting for the whole thing to finish before seeing anything.
- Any issues discovered along the way.
One thing that does not appear until the run fully finishes: the
video replay. The video is stitched together from all of the run's
screenshots only after every step has executed (and, if the run navigated
successfully, after the accessibility scan runs too), so during running
status you'll see individual step screenshots populating live, but the
combined video player only appears once the run reaches completed or
failed.
As soon as the run reaches a terminal status, the same polling refresh also brings in the Re-run, Edit & Re-run, Download, and View shareable report buttons automatically. You don't need to leave the page and come back to see them appear. See Editing & Re-running a Test for the difference between Re-run and Edit & Re-run.
Step results
Each step in the list shows:
- Its number and action type (
navigate,click,type,wait,assert). - Its target description.
- A 🤖 AI badge if that step's element was resolved by Claude Haiku 4.5 specifically, rather than the deterministic fallback matcher. See Step Types & AI Resolution.
- Its status:
passedorfailed(steps that haven't executed yet, in a still-running scenario, show no status badge at all). - If it failed, the exact error message. This is almost always
could not find an element matching "<description>"for a resolution failure, though a step can also fail for other execution reasons (a click or type action erroring out against the browser itself). - A screenshot taken immediately after the step executed.
Issues found
If the run turned up any issues (a failed step, a slow step, browser console errors, or accessibility violations), an Issues found section appears on this page too, above the step list, with a severity badge and category on each one. This is the same issue list that also appears on the run's public shareable report. See Issues & Accessibility Scanning for exactly how each category of issue gets generated, and Shareable Reports for the public report page itself.