As suggested on https://codereview.qt-project.org/c/qt/qtbase/+/328261 , it would be beneficial to set up automated (Squish-based?) runs of things like Qt Designer, Qt Creator, Assistant and so on. The point is to smoke test Qt for patches that otherwise pass "normal" tests, but may trigger regressions in end-user apps (also showing perhaps that making the patch in the first place was "wrong" due to some unexpected circumstance).

Bonus if these runs are actually done under a ASAN/UBSAN/TSAN-enabled build, to catch even more problems. See also QTQAINFRA-4139