Tom Frazier, an AI enthusiast and business consultant, has released Slop Mop, an open-source Chrome extension that flags low-value posts on LinkedIn. The tool uses Jev, a probabilistic decision-making AI model from TypeSafe, to score posts based on nine signs of 'slop' identified by SEO firm Graphite Growth. Frazier says the extension is free, requires no signup, and is designed to catch poor human writing as well as AI-generated content.
Slop Mop's nine tells include hype words, engagement bait, flowery language, and stiff wording. To avoid penalising quality content, two additional questions lower the slop score if a post sounds like a particular person or contains something useful to readers. A 12th question about whether a model drafted the post can only reduce the score for writing that otherwise reads as human; it cannot trigger a flag on its own. Frazier describes the extension as a filter for internet garbage rather than an AI detector.
The extension runs on Frazier's own Jev API account, so he has implemented a daily cap of 250 analysed posts per installation to keep costs down. He told The Register that early usage costs about $0.005 per user per day. Frazier says the project is primarily an experiment with Jev, which he calls 'a new class of AI infrastructure' that could do more for enterprise workflows than LLMs. If it proves useful and affordable, he may extend Slop Mop beyond LinkedIn.