It's a !ml flag, so there's a high likelihood it's wrong, because there's no downside to false positives. (The stuff I work on, which is all open source and do

It's a !ml flag, so there's a high likelihood it's wrong, because there's no downside to false positives. (The stuff I work on, which is all open source and doesn't even try anything silly, routinely but randomly gets flagged for some reason.)