I explain the details about it on today's high tea . I built a Claude skill that turns a process, an idea, or a messy folder into an ICM workspace. The folder structure does the orchestration. Numbered folders carry the order, the hierarchy scopes context, and plain markdown files hold state. One agent walks the right files at the right time and does the work a multi-agent setup would. Repo: github.com/RinDig/icm-architect 📦 What it does Two modes. 🔨 Build. You describe your work and it pulls out the structure already sitting in how you talk about it. The stages, the points where you stop and check, what stays the same every run versus what is new. Then it picks one of five proven forms and scaffolds the smallest workspace that carries the job. ♻️ Restructure. Point it at a folder, repo, or vault you already have. It reads every file, sorts each one by role, shows you a migration map, waits for your yes, then moves and checks the result. 🧩 The five forms Pipeline, umbrella, record library, knowledge bundle, context map. They mix and nest, so most real workspaces use more than one. ✅ The walk test Every result gets checked cold. An agent with no memory has to open the root, find its way, act, and report status from the files alone. If it can't, the structure gets fixed until it can. ⚙️ How to use it You can honestly just tell claude to download it from the link, but if you're using codex or something else it will just have to restructure the claude.md to agents.md Or if you want to do more Hands-On install Claude Code: drop the folder in ~/.claude/skills/icm-architect/, then say "ICM this" or "build me a workspace for X." Claude apps: zip the folder and upload it under Settings, then Capabilities. Fork it, break it, tell me what you built. 👇

Last month I was speaking with a friend who works for an engineering firm in Australia about ai, and all the cool things we can do with it these days, and he mentioned that he was trying to push to get a monthly newsletter out to their team to inform them of upcoming professional development courses and workshops. Of course it takes a lot of time to manually search the relevant websites, and put together a newsletter, etc. so no one has done it. I asked a few more questions about the tools they use, and then went and built out a small, structured ICM folder system, with the exact same blueprint that we have been learning in here and using for the competition building. I then made a loom video showing how it worked, and then emailed it to him along with the fully company branded email that it output. It was near the end of financial year at the time, so they were a bit busy, and he said he'd get back to me. Today, 1 month later, he came back and accepted my quote of $600, which includes 2 rounds of revision. To get the draft up and tested, it took me probably about 4 hours, and then there will be another few hrs in finalising it (with the revisions). In reality I probably undersold myself, but this is a side project at the moment (I run a cafe, not an ai consulting business...yet!), and I was excited to have the opportunity at a real client to test against. Let me be clear, I'm not selling a fancy Ai loaded website or app or anything... It is literally 5 folders, with a top level claude.md file, instructing claude co-work (or claude ai or code - or any other Ai tool that can follow structured instructions with a renaming of the claude.md file) on exactly what to do. This is exactly what Jake has been teaching here, and it will stand the test of time. As Anthropic updates their interface, and Open Ai starts to take over claude or another better tool comes along, this ICM folder system will continue to do its thing. It may need some tweaks along the way, but it's not going to need it's whole codebase updated or anything, because it's all plain language, english instructions.

I wanted to make an anonymous shout out to someone I met in this group recently (I don't want to share the name as I don't want to overload him with help requests) He spent several hours with me recently to go over ICM. I'm new to both ICM and Claude and he graciously offered help. I reached out and within 24 hours was on a call with him and he helps me put together my first ICM structure and even offered to do another. I continued on what he showed me and have been on claude since 8am implementing what he helped me build and then started to add a new project. The new project was done in less than 10 minutes and works flawlessly. Advice: Connect with people, there are lots of very smart people here who are also very kind and are willing to help strangers. Additional shout out to @Jake Van Clief for being an example as he gives so much valuable info

Writing this one from the road. I'm a couple of days into a tour with my family, and I keep landing on the same thought I want to share with you all. The air out here. The trees, the leaves. The kids finding themselves again away from their screens. And the funny part is that the itinerary that made this walk through the woods possible, the one that lined everything up so neatly, came out of the system we've been building. The AI put it together so we could close the laptop and actually go be in it. That's the balance on my mind. Building and creating, then going out into the world to see how it gets manifested in real life. I've loved going back and forth with a lot of you lately about setting up the gateways, getting the rules right so the output comes out the way you want it later. Good conversations. And they keep reminding me how much the why matters. When you're clear on why you're building something, the parameters and the guardrails start to make sense on their own. You stop fighting them. So smell the flowers. Build the kind of things that make the flowers shine, that make the experience worth stepping away from the computer for. I'll be hopping in here when I can over the next two or three weeks between stops. To everyone who reads, replies, and engages here: thank you, from the bottom of my heart. The people who inspired me to put things out there, you're doing the same for all of us in this room. So let's keep cheering each other on, and let's be an inspiration to each other and to ourselves.

Jeetu Patel, Cisco's president and chief product officer, posted this after their keynote week: https://www.linkedin.com/posts/jeetupatel_we-are-at-the-beginning-of-one-of-the-most-activity-7469585414857687040-4mtk/ Most of it is silicon and platform news. One line is worth pulling out if you build ICM systems: "every agent action will be a routing challenge, a trust decision and a telemetry event." A $200B infrastructure company just said what's slowing agents down is a trust problem, not a capability one. Agents don't get adopted until something can vouch for what they produce. The whole field, top to bottom, is arriving at the problem the people in this community have been working on since day one. What I want to flag is the crack in his own triad. Two of the three are infrastructure. Routing and telemetry get solved from outside the agent, and Cisco will happily sell you both. The trust decision is different. Their word for it is observability, meaning behavior monitoring, watch what the agent did. But watching what an agent did tells you nothing about whether its output should have shipped. You can log every step and still send a confident wrong answer out the door. That gap is the opening for every ICM builder here. Observability makes a system readable. It doesn't verify anything. The part that holds the line is a gate at the point of action with the authority to refuse, and that gate depends on knowing what "correct" means for your specific work. Which is exactly what a platform can't template. Cisco can build the dashboard. It can't decide, for your build, whether this particular output is honest. That decision stays judgment, and judgment doesn't commoditize when the infrastructure gets cheap. So here's the possibility for this community: the biggest players are going to make agents observable. Very few are going to make them refuse. If your ICM system already has a real gate, a step that can say no and mean it, you're building the layer they just named as the frontier, at the one altitude a platform struggles to reach.