Yes! I’ve had to do this at a previous company, while leadership was demoing at a conference, there was a bug that would crash the system. Connected remotely through a debugger and listened for the condition, modifying state to keep the crash from happening. Can’t remember the details, but it was exciting.