FAQ & Known Limitations
A collection of specific questions and edge cases that don't fit neatly into one of the other pages, gathered in one place.
My step failed with "could not find an element matching...", but the element is definitely on the page. Why?
The two most common causes, in order of likelihood:
- The element only exists for a logged-in user, and nothing earlier in the run actually logged in. See Account & Sessions.
- The description names an action rather than an element. "Scroll to the pricing section," for instance, describes something you'd do with a mouse, not a clickable thing on the page. See Creating a Playtest Run.
Less commonly: the element is inside a cross-origin iframe (common for payment forms; see the caution in Scenario Templates), which VeriWasp's browser automation cannot reach into at all, regardless of how the description is worded.
Why does my navigate step's description not do anything?
By design. See Step Types & AI Resolution.
navigate always reloads the run's original Target URL and nothing else;
its description field is read-only in the step editor specifically to make
this obvious up front, rather than letting you write something misleading
there.
Can I test a localhost URL?
Only if VeriWasp's own server can reach it over the network, which a
localhost address on your personal machine cannot be, from a hosted
VeriWasp instance. The target URL needs to be reachable from wherever
VeriWasp itself runs, not just from your own browser.
Does Chaos Mode affect a normal run's steps, or only add extra ones?
Both, indirectly. Your scripted steps still execute exactly as described.
Chaos Mode doesn't change what a click or type step targets, but it
throttles the network for the whole run (making every step take longer) and
injects extra click/navigation noise immediately after each click/type
step completes. See Chaos Mode for the full breakdown.
How long are screenshots and video kept?
14 days, after which the underlying media files are deleted from storage (the report page and its text content remain indefinitely). Download the report before then if you want to keep the visual evidence longer.
Does re-running or editing a test affect the original run's report?
No. Re-run, Edit & Re-run, and Edit steps all create a brand-new, independent run with its own ID, report link, video, and screenshots. The original run and its report are never modified or deleted. See Editing & Re-running a Test.
I misconfigured a run. Can I fix the steps instead of starting over?
Yes. A draft run has an Edit steps link next to Run this test; a
completed/failed run has an Edit & Re-run link next to Re-run.
Both take you to the New Playtest form pre-filled with that run's target
URL, viewport, Chaos Mode setting, and every step, ready to edit before
creating a fresh run from them. No credit is spent until you click Run
this test on the result. See
Editing & Re-running a Test.
Can someone else see my runs if I don't share the link?
No. The My Runs list and the private /runs/{id} detail page are only
visible to your own logged-in account. Only a run's specific, unguessable
public share link (/r/{slug}) is viewable without logging in, and only to
someone who actually has that exact link. See
Shareable Reports.
What happens if I run out of credits mid-workflow?
Starting a run (via the Run this test button, Re-run, or a
CI/CD trigger) checks your balance first. If it's
insufficient, nothing is deducted and nothing is queued. You'll see an
out of credits message (or, for a CI trigger, an HTTP 402 response)
pointing you to Billing & Credits to top up.
Is there a way to assert on specific text content, not just element presence?
Not currently. An assert step's only check is whether an element matching
its description exists on the page at all. There's no separate mechanism for
asserting a specific value, count, or exact text content beyond what's
implied by the description itself. See
Step Types & AI Resolution.