TL;DR
Jack Roberts walks through how to build a 'Claude Code Operating System' — a visual dashboard that centralizes all your AI models, memory systems, skills, costs, and usage data in one place. The system includes a self-improving 'dreaming' feature that automatically surfaces daily recommendations to save time and money based on your actual AI usage patterns.
Key points
1
The Claude Code OS is a unified dashboard that connects all AI models (Claude, ChatGPT, Gemini, Deepseek, Codex), memory systems (Obsidian, Pinecone, Supabase), and skills into one visual interface
2
A 'dreaming' feature runs daily — analyzing your last 7 days of AI conversations, clustering by intent, flagging repeated manual tasks, and generating 4 high-leverage improvement recommendations automatically
3
The system tracks real ROI: it calculates time saved per skill, subtracts your total AI subscription costs, and shows net return — useful for both personal use and client reporting
4
An onboarding wizard auto-detects all AI models and memory systems on your computer, asks for your hourly rate, and configures the dreaming engine to your preferences
5
The system identifies cost inefficiencies in real time — for example, flagging when you are paying Opus-tier prices for tasks that a cheaper model like Haiku could handle
Actionable insights
→
Set up the onboarding wizard to auto-detect your AI models, memory systems (Obsidian, Pinecone, Supabase), and input your hourly rate so the OS can calculate accurate time and money savings
→
Configure the dreaming engine with your 8 intelligence dimensions (conversation analysis, cost intelligence, skill performance, memory health, session hygiene, workflow patterns, external opportunities, business context) and set a daily run schedule
→
Use the skills tracker to identify unused or outdated skills to remove, and let the conversation clustering flag any repeated manual tasks that should be turned into automated skills
→
When onboarding clients, use the ROI calculator inside the OS to show them exactly how much time and money their AI skills are saving — making the value of your work concrete and measurable
→
Monitor the cost intelligence layer to catch model mismatches — if simple tasks are being routed to expensive models like Opus, the system will recommend cheaper alternatives like Haiku
Notable quotes
“Every single day, whether you touch it with your hands like this, whether you're on the beach sipping margaritas, this finds improvements for you automatically.”
“We don't wait for Anthropic to release anything. We just go ahead and start building it.”
“You're paying Opus prices for jobs that Haiku can do.”
Worth watching?
⏭️
Worth watching the full video?
The core concepts and setup steps are all captured here — watch only if you want to see the live dashboard UI walkthrough and the onboarding wizard in action.
Saved you some time? The creator still deserves a like.
Found this by searching? That's the hard way.
summree watches the AI channels you follow and sends you the summary the moment a video drops. Nothing gets past you. From £4/month.
Try it free