First met in aicp-c3-session-13 · Productivity & source apps
Local-first markdown notes with a knowledge graph — the course's second-brain substrate, pointed at by Karpathy's LLM-wiki writeup.
You want an AI-readable personal knowledge base that lives in plain files you own.
Plain markdown on disk: agents read and write it natively
Graph view turns linked notes into a navigable second brain
Notes, backlinks, graph visualization, plugin ecosystem (incl. LLM wiki plugins)
Doesn't sync or collaborate by default — local-first is the point
The storage layer under personal-assistant/second-brain builds (Devesh's Jarvis project).
Productivity & source apps
The KB's file conventions already fill this role for course knowledge.
current
notebooklm: hosted digestion vs owned files
| Session | Coverage | Moment | What happened |
|---|---|---|---|
| Office Hours III: Backing Up Claude Chats with Obsidian, Bridging Chat and Cowork, What Fable Is For, and the Context-Rot Problem | explained | Local second brain for Claude chats and outputs | |
| Office Hours IV: PostHog over Plausible, the Launch-Fast Loop, Scraping Ethics, Agent Swarms with Memory, and Why You Never Give Credits Away Free | mentioned | Knowledge graph, not agent memory | |
| Session 13: Claude Code — Skills, Sub Agents & the Token Economy (posted as 'Beyond Vibe Coding — n8n & More') | mentioned | 1:22:26 | His in-progress second brain, via Karpathy's LLM-wiki blog |
| Understanding Jev, the AI Built for Decisions (Sprint Day 2): Two Second-Brain Systems - a Leadership Self-Review Built Problem-First, and a Tag-Classifying Knowledge Base Where a Cheap Decision Model Replaces a Chat Model | mentioned |