I see this question a lot, so I wanted to share my take. ICM (the stage-by-stage folder setup, where each stage loads only the context it needs) cuts waste. It doesn't cap usage on its own. If your usage is still high, it's usually one of these: **1. Your context files are too big.** This is a symptom of ai generated slop without proper human auditing, maintain your files! If each stage loads a 5,000-word reference doc "just in case," you've moved the bloat into a folder. Check what each stage actually reads. Keep stage instructions short and push reference material into files that are only loaded when needed. **2. Stages aren't isolated.** ICM only saves tokens if each stage starts clean. If you run everything in one long conversation, the model re-reads the whole history on every turn. Start a fresh session per stage and pass work forward through output files, not chat history. **3. Too much gets loaded at the start.** A bloated root `CLAUDE.md`, or a pile of always-on rules, skills and MCP tools, is paid for on every single message. Trim the always-on layer to the essentials. Manage your hooks and permissions properly! **4. Outputs are being re-read in full.** If stage 3 reads the entire raw output of stages 1 and 2, you're paying for it again. Have each stage write a short summary or handoff file, and have the next stage read that instead. My /handoff skill is one of my favorite and most used. One of the few things that's persisted in all my "restarts". **5. Your model choice is off.** Running the strongest model on every stage burns usage fast. Use a lighter model for mechanical stages (formatting, extraction, file moves) and keep the heavy model for the stages that need judgment. !! I was just burned on this today. I had my fancy new process all shiny and polished. Yesterday I was polishing and left a mark - today, Opus and Fable CRUSHED my account usage before I could notice, and I feel like I'm a careful person. **6. Loops and retries.**

Hi all 👋 This is my first post and video. I have been a VIP member since March. I continue to learn and get value from all aspects of this community on a regular basis. I also joined the Lyceum recently to better understand how to translate everything that I have been learning and processing for the last ~6 months into a tangible skill, product, business and more. The progress that I have made so far has been well worth it. Given that its structured like a class, there are assignments, deadlines, group work, etc. That means more opportunity to procrastinate! In my most recent bender of procrastinating, I decided to be productive and try new things. I made a video in attempt to summarize many recent updates to how I work with Claude on a day to day basis. So here's my version... and it was supposed to be 30 seconds. It's version five and a bit over five minutes. I got way off track, lost the plot and burned way too many tokens - but I learned a lot and fun along the way! Hope there is something in there for at least 1 person 🙌 My Human Harness - First video I've ever made - Posting my imperfections

Deep Thought gave the perfect answer, and nobody knew the question 😄 I think many of us build with AI the same way, we get fast answers but never wrote down the real question. Here is my stressed out Deep Thought for the mood. I'm curious: what is the one question u should have asked ur AI before u started building?

Congratulations to @Bryson Bossley , pulling insane numbers on the weekly leaderboard. One of our OG's here at Clief Notes. Since Bryson already has Lifetime VIP, he gets a 30 minute 1-on-1 with Jake! ⏰ The 7-day clock resets. Next Monday someone new takes the top spot. 🎯 How it works: - 📝 Post things worth reading - 💬 Help people in the comments - 🛠️ Share what you're building, what's working, what's breaking - ❤️ Engage with what others are putting out there The leaderboard tracks all of it.

I just want to thank @Jake Van Clief for this community. So, I just made my first animated video and it's not perfect but I love it the same way a kid loves their lovey even when it's dirty and damaged!! I animated a song I wrote yesterday as I was wondering if it was possible to be authentic on social media, in general, but mostly for me. I was thinking how winning just looks like passing for me. I truly do not like to follow the crowd but when it feels like that's your only option, well you get...me: Trying to make money doing what everyone else is doing, but getting paralyzed because it doesn't align with my principles and values. Not to mention the fear of vulnerability and all the other fears that are involved. Anyway, that brought me to writing these lyrics and using hookpad to do the piano and melody. Then I sent it over to Suno for a quick demo with the AI voice modeled after me. I had a vision for the video and this video comes close so I am pretty stoked by it. It's just a rough draft as I intend to sing the song myself, but I really wanted something to post in the community 😅. This is going to be a part of my YouTube studio automation which I began building out over the weekend. It starts from a seed which comes from my chats with both Claude and Chatgpt, which Claude puts into a file for me. Either I pick the seed/topic or Claude will based on my current journey and interests. It then gives me some keywords and does a keyword search using the Eden.so MCP and it uses my own reasoning tools to analyze the scripts from YouTube or wherever and finding my gap. Where I fit in based on who I am which Claude has access to stuff but I also gave it that md file on me. Anyway, once we determine my gap and it searches for anything I have already said on the subject and gets my interview. It gives me a template and then writes the script for me based on my voice profile and thr YouTube script analysis I have already had it do.