How do you stop tasks and conversations from repeating each other? I built a Kanban board on top of my context. Whoever's on shift, Claude Code, Codex, anyone, knows where we stand. No circles reading files.

I am the biggest fan of the ICM-architect and I have been running many projects hours on end in parallel that all started with building the ICM structure and routings via the ICM architect. Most of my projects are pretty similar either software, hardware/IoT or business and the architect offers robust routing and coordination of the projects. A few things I noticed: - Similar project but gets completely different ICM structure (still gets the job done) - New project assumes information off one running in parallel - Some run fully autonomous goal-based and others require constant YES-ing I dont want to go and read all md files anymore (we are lazy) What would you add to make it behave in the same "structured" way like how ICM was intended to be? Share your ideas

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

Last weekend @Jake Van Clief went over how he was using his new work flow on this large consulting gig he is doing. I saw the drop first and looked at it but wasn't sure how I could use it but AFTER our high-tea call, my understanding became: "It's interviewing 1,000s of employees and helping find the gaps in handoffs from human to human and ai to human...and where AI is needed and where it isn't." Then it generates it visually in a nice "presentation" style thing that he can share with the executives (or really anyone to show the results) That new workflow audit he dropped is here on this post I don't work with huge companies. I work with small business owners & operators. The big "AHA!" for me was the "interviewing" part was an "if this than that" logic I was trying to build on the front end of a software, and now I could take that part of his workflow audit and use it there. I was pretty excited. THEN, just a couple of days ago, one of my VIP members DM'ed with a CLUTTERED mess. Like Jake and I, this guy is also a Marine. And just like I did, this guy ran straight at the AI gunfire with several businesses and personal projects, creating a "copy and paste errand boy" mess that he has to re-teach daily. As explained here in this post that Jake made about my simple explainer when ICM "clicked" He had done what many of us had: Youtubed stuff, listened to a "guru" and duct taped a workflow together that burned tokens and had robots putting US to work for them. So, I took the interview part from Jake's audit and I added 3 more moves, handed it to @Mason Page (the Marine in question) and he's executed the 4 moves that's decluttered his mess. Then, I ran Jake's ICM audit skill on it (the same way I audit my stuff weekly) and it landed right in the middle 2/4. We're cleaning up those 2 things right now and then we can give a warm welcome to another human saved from the multi-agentic, copy-paste, daily reminders to a robot, crap that many other humans are stuck in.

Okay — I think we accidentally have a really good ICM experiment. Ox Alpha is getting a ridiculous amount of attention right now. I initially wasn't particularly interested in another "look how good this new model is at coding" conversation. Then something clicked. People are running it through agent environments like Claude Code. Which means we can separate the brain from the system. Most of what I've built with ICM has Claude sitting underneath it. And that leaves a question we've never really been able to answer: How much of what works is ICM — and how much is simply Claude being Claude? Then an insight popped into my head: 'let's test it.' Take one mature ICM worker. Same files. Same rules. Same examples. Same methodology. Same task. Run it with Claude. Then change the brain to Ox Alpha. And DON'T optimize the ICM for Ox. That's important. If we have to rewrite everything to make Ox understand it, we've learned something. If it completely falls apart, we've learned something. If it follows the methodology but produces noticeably different judgment, we've learned something. But here's the result that would really fascinate me: 'What if a completely different model enters the existing context and starts behaving recognizably like the same worker?' The model supplies cognitive capacity. But the methodology, decisions, standards, examples, boundaries and accumulated context are supplying something else. We've transferred part of the capability into the environment. Or, put another way: 'We changed the brain and the person still knew how to do the job.' I have absolutely no idea whether that's what will happen. Which is exactly why I think we should run it. And I think we should decide the test criteria before we do. What would you guys measure?