Slop Counter — AI Slop Report Log

Keep your own private log of the posts you report as AI slop: how many, which ones, when, and who wrote them.

Report a post as AI slop and it disappears. You get a thank-you card and nothing else — no count, no history, no way to look back at what you reported or who wrote it. Reload the page and even the card is gone. Slop Counter keeps your own record instead. An independent tool by snipKI. Not a LinkedIn product — see the note at the end. WHAT YOU GET • A count. Every post you report, on the toolbar icon and in the popup. • A library. Which post, when, who wrote it, and the first lines of the text. • A dashboard. Reports per day over 30 days, the authors you report most, and a score distribution. • Export. JSON or CSV, whenever you want. Clear all, whenever you want. HOW IT WORKS You keep reporting posts exactly the way you already do — open a post's "…" menu and pick the AI-slop option. Slop Counter notices your own click and writes a line to a local log. It changes nothing about the feed, adds no buttons, and reports nothing on your behalf. It is a notebook, not an automation. WHO GETS CREDITED If someone reshares a formulaic post into your feed, the entry credits the person who WROTE it — with "shared by …" underneath. Blaming the messenger was the first version's worst bug, and fixing it is why attribution gets its own code path. Quote-shares only score the quoted post, never the sharer's commentary. THE FORMULA SCORE — AND WHAT IT IS NOT Each logged post gets a "formula score": how closely it matches the stock engagement-post template. Em-dash density. "It's not X. It's Y." One-line paragraphs. Emoji bullets. Engagement-bait closers. Rule-of-three cadence. Nine weighted signals, English and German. It is deliberately NOT an AI detector, and the extension never claims a post was written by AI. AI-detection classifiers are unreliable, and their false positives land hardest on people writing in a second language. Formulaic writing is what people are actually reacting to when they call something slop — and a human ghostwriter following a template scores high here, correctly. The score is context, not a verdict. It records what the detector would have said at the moment you reported something. NOTHING LEAVES YOUR BROWSER No account. No server. No analytics. No remote endpoint. Not even a webfont CDN — Poppins is bundled in the package, because a font request would leak every popup open to a third party. Your log lives in chrome.storage.local and goes away when you uninstall. The only permissions requested are local storage and access to the site where the feed lives. WHY KEEP YOUR OWN LOG Reporting a post looks like a signal rather than a moderation action: the post is hidden from your view and no record is kept that you can go back and read. Whatever it does behind the scenes, the part you can see is a one-way door. So this keeps the part you can see. WHERE IT WORKS linkedin.com, in English and German feeds. Nothing else — no other site is requested and no other site is touched. HONEST ABOUT BREAKAGE This is built on a feature that is still rolling out. Menu labels get translated. Internal names get renamed. Class names are hashed and change on every deploy. Detection uses two independent paths so one can survive the other breaking — but both breaking at once is entirely possible. If the count stops moving, that is what happened, and an update will follow. NOT AFFILIATED WITH LINKEDIN Slop Counter is an independent browser extension made by snipKI. It is not affiliated with, endorsed by, sponsored by, or connected to LinkedIn Corporation in any way, and it is not part of the LinkedIn service. It contains no LinkedIn code, artwork, or branding. "LinkedIn" is a registered trademark of LinkedIn Corporation and appears here only to describe where the extension works.