Fluff Meter for LinkedIn

Fluff Meter rates every LinkedIn post by how much of it is empty words instead of facts, and folds the ones you don't want to read. ON EVERY POST • A Fluff Index from 0 to 100%. Nothing concrete and corporate jargon push it up. It's the only number. • Clichés, in orange: engagement bait, humblebrag, too-neat fables, truisms, hustle, broetry, reads like AI. • Good signs, in green: real numbers, a real take backed by facts, a mistake the author owns. • What the post is about: know-how, news, hiring, career moves, promo and more. Click any chip to see what it means, with an example. FOLD WHAT YOU DON'T WANT Pure fluff folds into one line, and one click opens it again. Two short lists do the rest: • Fold: too much fluff, categories you skip (promo, events…), clichés you can't stand, topics of your own like "crypto". • Always show: categories you want, topics like "Rust", posts with a good sign. Make a rule right from a post: click a chip and pick Fold or Always show. Or start from a preset: Engineer, Recruiter or Job seeker. YOUR FEED IN NUMBERS The popup shows how much of your feed was fluff today, this week or this month, the top clichés and how many posts it folded for you. HOW IT WORKS Scores come from Jev by TypeSafe, a fast model built for rating text: one call per post, about 10,000 posts for a dollar. It starts in demo mode with random numbers; add your own TypeSafe or OpenRouter key for real scores. It reads the text only, not images. Posts about a loss, a war or an illness never get a score. PRIVACY No servers, no account, no analytics. Your key, settings and stats stay on your device. Only the text of the posts you scroll past, and your own topics, go to the AI provider you picked. Open source: https://github.com/horbel/fluff-meter It rates posts, not people. Not affiliated with LinkedIn or TypeSafe.