Python's Hypothesis automatically saves the random seeds that find counterexamples, so the next time you run your test suite, you may or may not be able to reproduce that bug.