You now have soft assertions for web, iOS, Android, and Windows flows. A failed expect.soft records the failure, but allows the test to keep running. One run can tell you about every failed check in the flow, instead of stopping at the first one.
How it works
A hard assertion stops the test at the point of failure. Nothing after that point gets tested until the issue is resolved. Say a release has two separate bugs: the product page shows the wrong price for a mug AND the "products in cart" counter stops counting items.
As the flow runs, a hard assertion would fail the test when it saw the wrong price. You would have to fix the price bug, then re-run the test to find the counter bug. Making the price check a soft assertion would let the test run all the way through and catch both bugs in one run.
Good to know
QA Wolf's agent now uses expect.soft by default if nothing further in the flow depends on it, like a label, a total, or a message.
Checks that a later step relies on, like the page loading or the record existing, stay hard assertions.
At the end of the run, all the failures are itemized on the run result and flagged on the video replay.
AI Investigation diagnoses each failure on its own, so failures with different causes become separate bug reports instead of one.