Semantic Sentry

Advisory layer for semantic moderation issues on Reddit. Mods only.

Semantic Sentry is an AI-powered moderation assistant for Reddit moderators. It scans comments in real time using Google Gemini to detect policy violations, hate speech, harassment, and semantic issues that keyword filters miss. Key features: • AI scan verdicts with confidence scores and recommended actions (remove, warn, escalate) • Image and GIF evidence analysis — detects violations in visual content • One-click escalation to the CollabGuard mod queue for team review • Mod verification — confirms you are a moderator before enabling actions • Works on any subreddit you moderate, no setup required • Private scan results stored per-subreddit, never shared across communities