How to Build Your Second Brain

@karpathy dropped a post describing how he uses AI to build personal knowledge bases. The idea is simple: instead of keeping notes scattered across apps, you dump everything into one folder. Then you tell your AI to organize all of it into a personal wiki - summaries, connections, articles - that gets smarter every time you use it. No special software. No database. Just folders and text files. In under 7 minutes you'll learn: The exact folder structure to set up (takes 2 minutes) How to automate web scraping into your knowledge base with one CLI tool The one-file "schema" that makes the whole system work How to get your AI to compile raw notes into an organized wiki The compounding trick that makes it smarter every time you use it The health check that catches mistakes before they pile up Follow @NickSpisak_ for practitioner-level AI implementation and Claude Code architecture content. Want the skill that builds this system for you? Grab the free LLM Knowledge Base skill here: https://return-my-time.kit.com/286e11f7e6 Alright let's get started... 1. Create Three Folders (2 Minutes) Open your terminal or file explorer. Create a project folder anywhere on your computer. Inside it, create three subfolders: That's it. This is the same structure @karpathy uses. The raw/ folder is your junk drawer of source material. The wiki/ folder is where the AI turns that mess into something organized. The outputs/ folder is where answers to your questions live. No apps to install. No accounts to create. Three folders. 2. Fill Your Raw Folder (10 Minutes) This is where most people stall. They create the folders and then stare at an empty raw/ directory wondering what to put in it. The answer: everything. Copy-paste articles into .md or .txt files. Save screenshots or diagrams as images. Export notes from whatever app you use now. Paste in meeting notes, research papers, project docs. Dump in bookmarks you've been hoarding for months. Don't organize it. Don't rename anything. Don't clean it up. That's the AI's job. I keep 17 raw source files across my X content pipeline - clipped articles, competitor breakdowns, analytics reports. None of it is organized by hand. But Karpathy didn't mention the part that actually speeds this up: automated collection. 3. Automate Source Collection With Agent Browser (Optional But Powerful) Vercel Labs just shipped agent-browser - a free CLI tool that lets your AI agent control a real browser. 26K+ GitHub stars. Two commands to install: That second command downloads a dedicated Chrome browser. Now your AI can scrape any webpage, extract the text, and save it directly into your raw/ folder. Here's what that looks like in practice: That's it. The AI opens the page, grabs the article text, and you pipe it into a file in raw/. No manual copy-paste. No browser extensions. Agent-browser handles pages that copy-paste can't touch. JavaScript-heavy sites that load content dynamically. Pages behind logins. Research papers with interactive figures. Anything that requires scrolling, clicking "load more," or navigating through menus before the content appears. The tool uses 82% fewer tokens than Playwright MCP, which means your AI agent can scrape 5-6x more pages within the same session. I use it to pull competitor articles, trending threads, and research docs directly into my pipeline without ever opening a browser myself. For your knowledge base, the workflow is simple: find an article you want, tell your AI "scrape this URL and save it to raw/", and agent-browser handles the rest. Your raw/ folder fills itself. 4. Write Your Schema File (5 Minutes) This is the part most people will skip. Don't. Create a file in the root of your project called CLAUDE.md (or AGENTS.md or README.md - the name doesn't matter, the content does). This file tells your AI what the knowledge base is about and how to organize it. Here's a starter template you can copy right now: @karpathy confirmed he keeps his schema "super simple and flat" in an AGENTS.md file. No database. No plugin. Just a text file that tells the AI the rules. This is the equivalent of what I use CLAUDE.md for across every project. It's the AI's instruction manual for your specific knowledge base. 5. Tell Your AI to Compile the Wiki (15 Minutes) Open Claude Code (or Cursor, or any AI coding tool that can read your files). Point it at your project folder and say: "Read everything in raw/. Then compile a wiki in wiki/ following the rules in CLAUDE.md. Create an INDEX.md first, then create one .md file per major topic. Link related topics. Summarize every source." Then step away. Let it work. When it's done, you'll have a wiki/ folder full of organized articles - connections you didn't see, summaries of things you forgot you saved, and an index file that makes everything searchable in seconds. The important thing: you don't edit the wiki by hand. That's the AI's job. You read it, you ask questions against it, and the AI keeps it updated. 6. Ask Qu