SeungJu Lee seungjulee

LLM Wiki v2

A pattern for building personal knowledge bases using LLMs. Extended with lessons from building agentmemory 20K+ Stars ⭐️, a persistent memory engine for AI coding agents.

This builds on Andrej Karpathy's original LLM Wiki idea file. Everything in the original still applies. This document adds what we learned running the pattern in production: what breaks at scale, what's missing, and what separates a wiki that stays useful from one that rots.

What the original gets right

The core insight is correct: stop re-deriving, start compiling. RAG retrieves and forgets. A wiki accumulates and compounds. The three-layer architecture (raw sources, wiki, schema) works. The operations (ingest, query, lint) cover the basics. If you haven't read the original, start there.

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

PostgreSQL is Enough

See the new site: https://postgresisenough.dev

Next.js client-only SPA example

Made this example to show how to use Next.js router for a 100% SPA (no JS server) app.

You use Next.js router like normally, but don't define getStaticProps and such. Instead you do client-only fetching with swr, react-query, or similar methods.

You can generate HTML fallback for the page if there's something meaningful to show before you "know" the params. (Remember, HTML is static, so it can't respond to dynamic query. But it can be different per route.)

Don't like Next? Here's how to do the same in Gatsby.

Note

I’ve written a newer optimized 2026 version of this guide with a faster setup using Zinit + Starship and improved plugin loading.

This version avoids slow shell startups and works better with modern Zsh setups.

Updated guide: https://gist.github.com/n1snt/2cccc8aa5f7b645a7628d3512c70deb6

	"""
	The most atomic way to train and run inference for a GPT in pure, dependency-free Python.
	This file is the complete algorithm.
	Everything else is just efficiency.

	@karpathy
	"""

	import os # os.path.exists
	import math # math.log, math.exp

	# 2023-11-27 MIT LICENSE

	Here's the open source version of my ChatGPT game MonkeyIslandAmsterdam.com.

	It's an unofficial image+text-based adventure game edition of Monkey Island in Amsterdam, my home town.

	Please use it however you want. It'd be nice to see more ChatGPT-based games appear from this. If you get inspired by it, please link back to my X https://x.com/levelsio or this Gist so more people can do the same!

	Send me your ChatGPT text adventure game on X, I'd love to try it!

	type MessageAction struct {
	Type string `json:"type"`
	Payload MessagePayload `json:"payload"`
	}

	// MessagePayload ..
	type MessagePayload struct {
	Message MessageWithInfo `json:"message"`
	}

	/*
	Instagram API: view and backup direct messages from a web browser

	Since April 2020, Instagram has a web version to send and read direct messages so my Instagram scripts are not longer needed and I would not recommend using them unless you really need it, to avoid being banned
	(never happened to me with Instagram but WhatsApp is owned by Facebook also and they did it to users registering from an unofficial app like yowsup: https://github.com/tgalal/yowsup/commit/88b8ad9581fa22dac330ee3a05fec4e485dfa634#diff-b335630551682c19a781afebcf4d07bf978fb1f8ac04c6bf87428ed5106870f5)

	1) Log in on Instagram web version and go to your profile page
	(the home page will not work because it loads data when scrolling down and the direct messages will be displayed at the bottom of the page)

	2) Modify HTTP headers with a browser addon like Header Editor (https://addons.mozilla.org/en-US/firefox/addon/header-editor/)

Service	Type	Storage	Limitations
Amazon DynamoDB	⚠️ Proprietary NoSQL	25 GB	⚠️ Payment method required
Amazon RDS	⚠️ Proprietary RDBMS		⚠️ Only free for 1 year
Azure SQL Database	MS SQL Server		⚠️ Only free for 1 year
👉 Clever Cloud	PostgreSQL, MySQL, MongoDB, Redis	256 MB (PostgreSQL)	Max 5 connections (PostgreSQL)