Hi everyone, I've been wrapping my head around ICM and agents. I'm a non-technical user even though I have a bit of background of workflow process engineering, and I'm still trying to understand what we were taught, and thought it'd be good to clarify with the community. When we build folders and sub-folders, can I say that the sub-folders are actually the "agents" that handle that part of the work? So if we have a content creation workflow, and research is a sub-folder, it is actually a 'research agent'. Is this understanding correct? So technically I can run hermes, claude cowork or goose cli on this folder, and it will react the same way (from the context and whatever is in the folders), is this correct? So technically if there is no requirement to have a build, it will just run as-is on the folders with no py script or type-script files, right?

This post is not for the ICM pro, there will be no talk of gates, scripts, or orchestration! This is for the person just starting out! @Karli Rosario Yes, I mean you! (And anyone else who may just be starting out with ICM) Seriously, I'm glad you found ICM. Let me give you the simplest version of it I know. ICM is a system of structured folders. Yes, the same folders you have been using on a computer for most of your life. The ones you stored photos in, & pirated music from Napster and LimeWire. That's it. I will take you through the process below. When working with AI, a lot of people are doing this 👇 You take a long prompt, feed the entire thing to AI at the beginning of your interaction, and spend time going back and forth with AI trying to get the outcome you want. (I'm not coming for you Karli, you are exceptionally good at this, but ICM will make your outcomes exceptionally better!) What is different about AI and prompting with ICM 👇 You take that same really long prompt and instead of giving it to the AI all at once in the beginning, you break it into steps, and each step gets its own folder, each folder gets its own piece of your large prompt, just 1 step from it, and you ordered the folders by when the steps happen in the workflow. You got it? Good 😊 ❤️‍🔥 -------------------Still a bit unclear, let's bake a cake. 💡 Here's an analogy I have success with (I picked this up way back in my VB programming days): Imagine teaching AI to bake a birthday cake. 🎂 The way most people do it: 👇 One giant prompt. "Bake a cake, here's the recipe, the frosting technique, the decorating style, the candle placement..." Then they hit enter and wait. The AI is juggling 40 instructions at once, and by step 30 it's forgotten step 3. The ICM way: 👇 Break the prompt/workflow into steps. Each step gets a folder. The first folder is your first step. Then you point the AI at the first step, and the first step is 00-birthday-cake: (Point the AI just means giving access to the folders to the AI, through uploading or direct local access, don't worry about that now, let's keep building our cake.)

Just curious, does everyone write your md files from scratch? Or do you ask claude to help you write, by giving it information? And can I say that the recommendation is to have md files lesser than 150 lines?

Hi everyone, just curious, if you can only choose to use Opus 4.8 in one area, would you use it during the planning phase where you're coming up with the folder structure, or would you use it to code?

I just wanted to put this out there for a moment, especially since I am currently going through Jake’s course myself. Jake is sharing his content for free, and on top of that, he has made his full GitHub repository with the ICM protocol available under MIT License which means people can actually use it, build on it, adapt it, and even implement it into their own systems. I think that deserves appreciation. What stands out to me is that this is not just another AI hype thing. It feels like he is working on something much more foundational: a structure that helps people organize, route, and actually use AI in a serious way. And that‘s powerful, because with something like this, you do not have to be afraid that the big companies will make your work obsolete. It is almost the opposite, as he said himself before. The better the models and tools from the big companies become, the more valuable a solid system like this becomes, because you can plug those improvements into the structure and keep building on top of it. Of course, how to integrate everything properly is still the next big step. I am still unsure about some parts myself, but that is exactly where a community like this can become really valuable. To me, that is a much healthier way to look at AI. You can ignore a lot of the hype, stop chasing every shiny new tool, and focus on the infrastructure that actually gives you leverage. Thank you Jake for making this available. Sharing something like this openly gave me a clear path instead of getting lost in the noise. ———————————————————————————————— @Jake Van Clief @David Vogel