Hey everyone, gather round. This one matters. 👋 We've been watching something, and it's time to name it. 🧭 The ICM posts, the AI breakdowns, the promo threads: solid work. Real value. No fluff. But some of those posts have a quiet second job: funneling traffic to something the author gets paid for. 🚨 Let's be precise about the line: Someone shares a great external resource? Fine. That's the good stuff. Someone shares their own work with no money in it, a free repo or an open tool? Fine too. That's building in public. Someone links their own COMPENSATED property (affiliate, referral, lead capture) so they personally benefit? Different animal entirely. 🐍 That's not sharing. That's extraction with a costume on. And it costs us. Every time: 📉 Trust drops. 🤐 High-quality members post less. 🔍 The signal gets harder to find. 📉 The feed gets noisier and less worth showing up for, even when the writing itself looks good. ➖➖➖ 📏 THE RULE, NOW OFFICIAL Simple. And let's be clear about what it is NOT. 🔗 Sharing useful external resources like a YouTube video, a guide, a tool you love? TOTALLY FINE. Keep them coming. That's the good stuff. What's NOT allowed: 1️⃣ LINKING ANYTHING YOU PERSONALLY BENEFIT FROM: affiliate cuts, referral bounties, lead capture. If personal gain is the engine, it doesn't fly here. 2️⃣ POSTS THAT ONLY EXIST TO FUNNEL TRAFFIC TO THAT STUFF. If the value lives behind your link instead of in the post, that's an ad, not a contribution. This applies to posts and comments, everywhere except the Connection Hub. Wanna pitch or network? Take it to the Connection Hub, and keep your links in your PROFILE. People click your name. They find it. That's the clean path. Think you're the rare exception? Ask a moderator first (@David Vogel or @Shirsho Guha). ⚠️ Fair warning: the default is no. ➖➖➖ ⚖️ WHAT HAPPENS IF YOU CROSS THE LINE No drama. No mystery. Just math: ⚠️ FIRST TIME: a moderator reaches out, and you edit or remove the post yourself. ⚠️ SECOND TIME: the post comes down, and you get a formal warning.

More people catching into the context engineering train and finally realizing it’s not all about a big long prompt https://www.instagram.com/reel/DeXXTT0la6t/?mdxt=MXY5Mm1mMXFlc3k1NQ==

@Jake Van Clief posted a video where he had a bunch of cards stacking over another that was reviews he pulled from Google or something like that. I’m thinking it was Remotion or something along those lines? Does anybody recall it? I am digging through the classroom content trying to find it. A friend just asked if I knew how to do something like that and it might be just what my friend needs?

I don’t mean your single person business or your mom and pop shop where you’re the AI guy and it’s on one PC. This is for your employer with 20, 50, hundreds or thousands of workers on PCs. Have you deployed AI to the general workforce yet? What does that look like? Is it: - An entire workforce with access to chat but no coding harness - Users have coding access but restricted sandboxing to prevent leaks - Users have full capability with something like Claude code or codex with ability to edit files at will How does your company manage this so the average user doesn’t compromise systems or worse cause a security risk. Let me know what it looks like below.

Hi i am interested on the Harness layer that you are using