Subcategory: Test pass rate / flakiness
Score: 8 → 10 (target)
make test is flaky. During assessment the first invocation exited code 2 (pytest-xdist worker crash) mid-run; an immediate clean re-run passed 143/143. The suite spins up real background servers/threads and uses time.sleep(server_timeout) for readiness, which is inherently timing-sensitive under -n auto.
Suggested fix: mock the server/thread boundary where behavior (not integration) is under test, and replace fixed sleep-based readiness with polling/health-check waits; ensure per-worker port isolation.
Done when: make test passes 10 consecutive runs with no worker crashes.
Subcategory: Test pass rate / flakiness
Score: 8 → 10 (target)
make testis flaky. During assessment the first invocation exited code 2 (pytest-xdist worker crash) mid-run; an immediate clean re-run passed 143/143. The suite spins up real background servers/threads and usestime.sleep(server_timeout)for readiness, which is inherently timing-sensitive under-n auto.Suggested fix: mock the server/thread boundary where behavior (not integration) is under test, and replace fixed
sleep-based readiness with polling/health-check waits; ensure per-worker port isolation.Done when:
make testpasses 10 consecutive runs with no worker crashes.