Nick Brown nick-brown

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.

My Elixir Study Notes

Important Links

Writing Assertive Code With Elixir

Notes

Get some help:

iex> i 'hello'
Term

Basics

engine/server (daemon)
- docker version or docker info
Image: application we want to run
- hub.docker.com (default image registry)
Container: the instance of that image running as a process
- many containers can run off of the same image

Regular Event Loop

This shows the execution order given JavaScript's Call Stack, Event Loop, and any asynchronous APIs provided in the JS execution environment (in this example; Web APIs in a Browser environment)

Given the code

	function subMatches(string, subs) {
	return []
	.concat.apply([], subs.map(sub => matchIndexes(string, sub)))
	.reduce((acc, idx) => {
	acc[idx] = true;
	return acc;
	}, Array.from(string).fill(false));
	}

	function matchIndexes(string, sub, startIndex=0, matches=[]) {