JevGuard

jev-guard checks every chat message with TypeSafe's Jev model and flags harassment, scams and grooming in about 200 ms. You decide what gets hidden, warned or sent to staff.