LARP Detector for LinkedIn

Marks LinkedIn posts REAL or LARP as you scroll. Flags humblebrags, engagement bait, and AI slop. Powered by Jev (TypeSafe).

LinkedIn is full of posts performing a persona instead of sharing anything real: humblebrags, "agree?" engagement bait, fabricated parables about baristas teaching leadership, and AI-generated filler. LARP Detector marks each post REAL or LARP as it scrolls into view, so you can skip the theatre and read the substance. HOW IT WORKS Each post is judged by Jev (TypeSafe System One) against a set of specific, observable questions: Does the author show work they actually did? Is the story checkable? Is the boast disguised as modesty? Is the ending engineered for reach? The answers are combined into a verdict, and hovering the badge shows exactly why a post was flagged. WHAT IT CATCHES • Humblebrags: boasts dressed as complaints or humility • Engagement bait: "Comment YES for the template", "Repost if…" • AI slop: template posts with no names, specifics, or voice • Fabricated stories: vivid parables with nothing anyone could check • Costume headlines: "Visionary | Top Voice" over a post with no substance • Name-dropping: status borrowed from famous people and companies WHAT IT LEAVES ALONE Plain job updates, layoff and hardship posts, requests for help, satire, and shared news are recognised and not flagged. YOUR CONTROLS • A slider sets how strong a post's LARP signal must be before a badge appears • Optionally show REAL badges too • Sponsored posts are labelled without being analysed SETUP Works out of the box in a free offline mode using built-in heuristics. For full analysis, add a TypeSafe or OpenRouter API key in Options. Your key stays in your browser and is never sent to LinkedIn or to the page. Github if any issues: https://github.com/sxeptical/larp-detector