OutputClassifies the toxicity level of a social media comment and provides confidence, justification, and recommended moderation action.