ThreadSense for Facebook

Flags likely trolling or harassment signals on Facebook using local heuristic analysis. Never makes definitive identity claims.

**ThreadSense for Facebook** helps you spot risky comment patterns while you browse Facebook, directly on the page, in real time, and privately on your device. It analyzes comments already visible to you and shows separate risk badges for: - Rude tone: insults, shouting, aggressive punctuation, and condescension - Troll pattern: repetition, copy-paste behavior, and multi-target hostility - Scam / bait: scam language, bait phrasing, and coordinated talking points - Fake-account risk: link-spam patterns, weak profile-style signals, and burst activity ## What it does - View per-comment risk badges - Open a detailed explanation popover for each comment - Use soft-hide modes to dim or collapse Medium and High risk comments - Snooze authors for 24 hours, 7 days, or always - Add custom block and allow phrases - Use optional regex patterns for advanced filtering - Export and import phrase patterns as JSON - Enable optional image-comment analysis and local OCR (experimental) - Turn on focus guard prompts during heated scrolling sessions - See quick tab stats and controls from the toolbar popup ## Important note - ThreadSense provides heuristic estimates only - It does not make definitive claims about a person’s identity, intent, or authenticity ## Privacy - Runs locally in your browser - No Facebook comment text is sent to ThreadSense servers - No external analytics or tracking scripts - Settings are stored locally on your device