Kris Ciccarello kmctown

LLM Wiki v2

A pattern for building personal knowledge bases using LLMs. Extended with lessons from building agentmemory, 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.

Remote debugging in Docker (for Django apps)

In order to enable debugging for your Django app running in a Docker container, follow these steps using Visual Studio (Code):

Add ptvsd to your requirements.txt file

ptvsd == 4.3.2

To your launch.json, add this:

My notes for Dokku on Digital Ocean.

These may be a bit outdated: Since I originally wrote them, I've reinstalled on a newer Dokku and may not have updated every section below.

Commands

Install dokku-cli (gem install dokku-cli) for a more Heroku-like CLI experience (dokku config:set FOO=bar).

# List/run commands when not on Dokku server (assuming a "henroku" ~/.ssh/config alias)

ssh henroku dokku

	require 'dnsimple'
	require 'platform-api'

	namespace :staging do
	desc "create subdomain DNS record for Heroku review app"
	task :publish_dns do
	heroku_app_name = ENV['HEROKU_APP_NAME']
	heroku_app_name =~ /.*(pr-\d+)/

	subdomain = $1

	//
	// README:
	// - Listens for PUSH events
	// - Fetches the ref pushed via the given remote
	// - Sets the repositories HEAD to latest ref
	// - Checks out the new HEAD (--force)
	// - Install dependencies from package.json
	// - Calls `npm run reload` (My app uses this)
	// - Calls `nginx -s reload` (My app also uses this)
	//

	// Restify Server CheatSheet.
	// More about the API: http://mcavage.me/node-restify/#server-api
	// Install restify with npm install restify


	// 1.1. Creating a Server.
	// http://mcavage.me/node-restify/#Creating-a-Server


	var restify = require('restify');

	#!/bin/bash

	# Source: http://blog.nonuby.com/blog/2012/07/05/copying-env-vars-from-one-heroku-app-to-another/

	set -e

	sourceApp="$1"
	targetApp="$2"

	while read key value; do