Live LabsBuildersKnowledge2 hours

Build a second brain: an LLM wiki that maintains itself

Instead of searching your notes every time, have a model write and maintain a wiki from them. We follow Karpathy's LLM wiki pattern and run it on our own sources.

You leave with

A markdown wiki that an LLM updates from your raw sources, with rules you control.

How the time is spent

  1. I do30 minParam builds it live

    I ingest three sources and show the wiki pages and links the model writes.

  2. We do48 minTogether, step by step

    We write the rules file that decides what the model may change.

  3. You do42 minOn your own code

    You feed it your own notes and review its first pages.

What you will be able to do

Synthesise once

Know when a maintained wiki beats search every time.

Set the rules

Control what the model writes and links.

Keep it honest

Trace every page back to its source.

Starts from

Before you come

For
Engineers and leads drowning in notes, docs and bookmarks.
You need
A coding agent you can run locally. Your own notes folder.
Stack
Claude Code, Markdown, Obsidian, mem0
Track
Hands-on agentic AI for engineers who ship.

More builders labs