The Blueprint
Everything I've learned is free. Here's the practical guide — the documents, the stack, the steps, and the traps.
Part 1
These six documents are the skeleton. Without them, you have disconnected AI conversations. With them, every AI session starts from current truth and every change gets logged. Skip any of these and the system drifts without you noticing.
The single entry point every AI reads first. It says what exists, what matters, and where to find everything.
How to create one: One page. List every venture, project, and life area. Link to each North Star. Update it only when the map changes.
Skip it and: Every AI session starts blind. You spend the first 10 minutes re-explaining your world.
Dense strategy docs — one per domain. These are constitutions, not daily notes. They tell AI what's true about a domain and what you're trying to achieve.
How to create one: Record yourself talking about the domain for 20 minutes. Transcribe. Ask AI to find the patterns. Refine into 2-3 pages. Update only when something real shifts.
Skip it and: AI makes assumptions about your priorities. You get generic advice instead of advice that knows your context.
What's built, what's running, what version it's on. The technical truth about the entire ecosystem. Any AI reads this and knows the current state.
How to create one: List every automation, script, integration, and tool. What it does, when it runs, what it depends on. Update after every build session.
Skip it and: You build duplicates. AI suggests things that already exist. You lose track of what's running.
Every significant change, logged with date and context. The institutional memory of the system. Any AI can pick up exactly where the last one left off.
How to create one: After every build session, log what changed, why, and what to watch for. Three sentences minimum.
Skip it and: Next session starts from scratch. You repeat work. You forget why you made decisions.
Specs that pass context from thinking to building. The paper trail between deciding and shipping. Thinking AI writes it, building AI reads it.
How to create one: After a strategy session, write: what to build, why, how it connects, definition of done. The building AI reads only this.
Skip it and: Thinking and building blur together. You get messy code that doesn't match the strategy.
Where the system surfaces things that need human judgment. Any part of the system can flag something. You review and decide.
How to create one: A simple list. Any AI or automation can append to it. You review daily. Decide, defer, or delete.
Skip it and: The system drifts on its own. Small wrong assumptions compound. You don't notice until something breaks.
Part 2
These are the six roles your stack needs to fill. I'll tell you what I use, but the role matters more than the brand. If a better tool shows up tomorrow, swap it. Your documents are the memory — the tools are replaceable.
Must be read/write, not read-only. AI needs to update pages, log changes, and create documents. This is where everything lives.
Examples: Notion, Obsidian
One place for all files. Root level equals truth. Everything AI touches is here. Where a file lives tells AI how much to trust it.
Examples: Google Drive, OneDrive, Dropbox
For strategy, not building. Load it with your North Stars and think together. Produces specs, not code. Keep this separate from your builder.
Examples: Claude, ChatGPT, Gemini
Reads specs, ships automation. Doesn't need your strategy context — just the handoff page and technical specs. Test fast, iterate, close out.
Examples: Claude Code, Cursor, Windsurf
Pushes what matters to you. If it requires you to remember to check something, it will fail. Everything pushed, nothing pulled.
Examples: Dashboard, Slack bot, daily page
Your builder writes it, your tasks run it. Automations, data sync, scheduled processes. The glue between everything else.
Examples: Python
Part 3
The principles took me three years. The infrastructure came fast once the foundation was in place. Here's the order I wish I'd done it in.
Not about your system — about your life. What matters to you, what you're working on, what keeps you up at night. Transcribe it. Upload to AI: "find the patterns I'm not seeing." Two pages max. That's your seed.
Why it matters: AI can find patterns in your thinking that you can't see yourself. This becomes the foundation for everything — your Psychological Blueprint, your North Stars, your entire system's understanding of you.
One dense document per major domain of your life. These are constitutions — they tell AI what's true about a domain and what you're trying to achieve. Updated only when something real changes.
Why it matters: Without these, AI makes assumptions about your priorities. With them, every conversation starts from truth instead of guesswork.
Root level equals truth. If you skip this, everything you build processes garbage. Where a file lives tells AI how much to trust it. Clean, verified, current files at root. Everything else in subfolders where AI knows to treat it as potentially noisy.
Why it matters: AI confidently cited a brainstorm I'd done two years ago as if it were my current strategy. The output sounded great. It was completely wrong.
Choose a hub that AI can write to — not just read. Turn off built-in memory on every AI platform. Your ecosystem IS the memory now. At the end of every significant conversation, you and AI decide together what's worth keeping.
Why it matters: When the platform decides what gets stored and when it shows up, you don't control it. My political views showed up in my tech conversations. That's noise.
Use one AI for strategy (loaded with your North Stars and documents) and a different one for building (loaded with technical specs and handoff pages). They coordinate through documents, not shared context.
Why it matters: Tools where you just talk to AI and it figures everything out are a disaster. No separation between thinking and building. You end up with a mess.
Don't design the whole system. Build one automation that saves you real time. Log what you built. Use it. Notice what's missing. Build the next thing. The flywheel starts turning.
Why it matters: The loop — use, strategize, build — is the compounding advantage. Every revolution, AI does more and you operate higher. But only if you actually use what you build.
Part 4
I fall into all of them. My AI catches them too — because it knows my patterns, it can tell me when I'm doing it. Here are the three biggest ways people sabotage their own AI systems.
Designing elaborate systems instead of shipping something that works. You spend three weeks on architecture and never ship a single automation. The system exists in your head, not in production. Ship something small. Make it real. Then iterate.
Starting twelve things at once because everything feels urgent. AI makes it easy to start things. Finishing them requires focus. Pick one domain. Build one automation. Get it working. Then expand. Parallel starts kill sequential progress.
Organizing instead of executing. You spend all your time tagging files, structuring databases, designing templates — and none of it produces output. Organization is a tool, not a goal. If you're organizing more than you're shipping, you're hiding from the hard work.
I'm documenting every step of building this — philosophy, build logs, practical guides. Free.