I’m building a reusable workflow kit from what I’m learning here: a task card, verification checklist and restart note that keep the goal, source material, results and next step connected. The aim is less repeated explaining and fewer “where were we?” moments. It’s an early template kit, not validated software. Sharing it for others to try and improve. Thanks to Jake and the community for the inspiration.

RinDig/icm-architect It runs in pretty much any agent now: Codex, Gemini CLI, Hermes, Claude Code, Cursor. Or hand the folder to any model and tell it to use the skill. Workspaces now start from an AGENTS .md file, so they aren’t tied to one company’s tools. It would work with others before just not as reliably, should be better now. I also cleaned it up a lot. The whole method comes down to three building blocks now: a sequence of steps, a shelf of same-shaped files, and a small map that points to everything. Every step ends at a check, whether that’s you looking at it or a small model doing it, so a mistake gets caught before the next step builds on it. There are also some small scripts that check your links and draw your whole workspace as a graph you can click through (very basic graph, obsidian is better for viz ill make a nicer one later) github.com/RinDig/icm-architect Install steps for each agent are in the README. Try it on something you do every week and tell me where it breaks.

My friend sent me a video of a runway show he saw. I asked my video worker to make a 45 second run way show with a background, then I also asked him to just make the stage with a background scene, as if I was showing what we could pay in the background of the show. I am kind of taken back by the results from just that request I gave the worker. What do you think?

I am going through the Roadmap in Decster, looking to create more, get more ideas... I came out with a different question... The material is what did it. The sessions come with files and a prompt at the bottom of each one. I ran them against my own work. The answers came back as questions, and the questions were hard. The hardest one was simple. Have you actually finished this for one person? I had not. I was designing a system for every trade and a template for students, and I had not finished it for one contractor. When I sat with that, I had to admit two things. I have about 20 hours a week and I was spreading them across everything. Building has never been my problem. Asking is. So I picked one client, a painting contractor whose purchase orders do not match what he measures on site. He did the check by hand to show me what it costs him. Now I know exactly what I am building against, and why. Focus fixed more than the build. With one client, the ask got small enough to make. What I would tell myself six months ago: you do not need another idea. You need to answer one question... Who is the one person you should finish this for, before you start anything new?

I’ve been developing my NeuroAdaptive Learning Environment—NALE—on and off for about three semesters. Along the way, the research, prototypes, revisions, and correspondence spread across four project folders, Google Drive, email, and my external SSD. There is some irony in becoming overwhelmed by the documentation for a project intended to make learning more manageable. 😂 So I’m applying ICM to that work, too. The goal is a shared map that helps me find the right material, understand how it developed, and pick up where I left off. Tonight’s concrete progress: we created a read-only inventory tool for my SSD. The first scan encountered a problem, we revised the tool, and the latest run indexed 71,058 entries, flagging 8,086 candidate files based on their names and locations. Those candidates still need content review; the count doesn’t tell us which documents are relevant or current. We’re keeping the original folders and historical versions intact. From there, we can connect confirmed records, trace decisions, and establish which versions actually represent the current work. I’m learning that preserving the history matters just as much as finding the latest file. An older draft may explain why a decision was made—or why we abandoned an approach. I’m still amazed that I can describe a problem this messy and collaborate with AI to build something practical for working through it. ICM is helping me give several semesters of work a structure I can return to and understand.