Of course the AI flagged you, it was built by the same kind of people who write those passive aggressive emails you're complaining about. Free tools like that are trained to spot "bad tone" which really just means whoever talks the most direct gets marked as the problem. The whole point of those sorters is to make everyone write in this fake polite corporate voice where you say "just circling back" instead of "you still owe me that report." If a machine thinks you're the toxic one, that might just mean you're the only person actually saying what needs to be said. Maybe the real issue is that everyone else is so scared of sounding rude that nothing ever gets done.