Unslop — AI slop filter for X

Unslop hides AI-generated replies and engagement bait on X before you read them, and it does the scoring inside your browser — nothing you read is ever sent anywhere. Replies on X are increasingly not written by people. Unslop reads each post in the page, scores it against a rulebook of slop tells, and collapses the ones that fail into a one-line bar you can expand. WHAT IT LOOKS FOR • Empty agreement and hollow compliments — "Absolutely brilliant insights!" • The "it's not X, it's Y" cadence that LLMs fall into • Corporate LLM diction — leverage, unlock, game-changer, delve • Comment bait — "Who else agrees?", "Drop a 👇", "Comment WORD and I'll DM you" • Listicle formatting used as filler • Bold-unicode obfuscation, folded back to plain text before scoring WHAT PROTECTS A REAL POST A filter that hides everything is useless, so human signals score posts back down: first-person action, hard numbers, code, admitted mistakes, real friction, informal register. A changelog is not slop for being a bulleted list. Measured against a corpus of 384 real X posts: 100% of labelled slop caught, 0% false positives on real replies. NOTHING IS DELETED A filtered post leaves a one-line bar with a Show button and the reason it was caught. Disagree often? Lower the strictness. PRIVACY The free version makes no network requests at all — it cannot phone home, and we have no way to know it is installed. Post text never leaves your machine, because there is no server that receives it. Full detail: https://useunslop.app/privacy FREE • Timeline filtering • Balanced strictness • On-device scoring PRO — $9 once, not a subscription • Reply filtering, where the slop actually lives • Strictness slider, 25 to 80 • Fade and label modes • Filter reasons on every caught post • 3 browsers, one key Google removed payments from the Chrome Web Store years ago, so Pro is bought at https://useunslop.app and unlocked with a key. 30-day refunds, no argument. Not affiliated with, endorsed by or connected to X Corp.