Kate Belisle akatie

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.

Chat GPT "DAN" (and other "Jailbreaks")

How to add a collapsible section in markdown

1. Example

Click me

Heading

Foo
Bar
- Baz

Qux

	"firm_name","entity_name","firm_slug"
	"1-800-FLOWERS","1-800-FLOWERS","1800FLOWERS"
	"1800FLOWERS.COM","1-800-FLOWERS","1800FLOWERSCOM"
	"1-800-FLOWERS.COM, INC.","1-800-FLOWERS","1800FLOWERSCOMINC"
	"1800 FLOWERS","1-800-FLOWERS","1800FLOWERS"
	"1-800-FLOWERS INC","1-800-FLOWERS","1800FLOWERSINC"
	"1800FLOWERS.COM,INC.","1-800-FLOWERS","1800FLOWERSCOMINC"
	ARMY AIRFORCE EXCHANGE SERVICE,AAFES,ARMYAIRFORCEEXCHANGESERVICE
	ARMY AIRFORCE EXCHANGE,AAFES,ARMYAIRFORCEEXCHANGE
	ARMY AIR FORCE EXCHANGE STORES,AAFES,ARMYAIRFORCEEXCHANGESTORES

	function [Class, BroadClass] = KoppenGeiger(T,P) %,classes)
	% [Class, BroadClass] = KoppenGeiger(T,P,classes)
	%
	% Classify monthly temperature and preciptiation climatologies according
	% to the Koppen-Geiger climate classification. The T and P inputs
	% represent the temperature and preciptiation climatologies,
	% respectively. T and P should be provided as three dimensional arrays
	% with the third dimension representing time (12 months). The temperature
	% data should have units degrees Celsius and the precipitation data units
	% mm/month.

	pip install streamlit
	pip install spacy
	python -m spacy download en_core_web_sm
	python -m spacy download en_core_web_md
	python -m spacy download de_core_news_sm

	<!DOCTYPE html>
	<html lang="en" style="background: black;">
	<head>
	<meta charset="UTF-8" />
	<meta name="viewport" content="width=device-width, initial-scale=1.0" />
	<meta http-equiv="X-UA-Compatible" content="ie=edge" />
	<title>Neon 3D bars demo</title>
	<style>
	#root {
	display: flex;

	##### READ SPARK DATAFRAME
	df = spark.read.option("header", "true").option("inferSchema", "true").csv(fname)
	# store the schema from the CSV w/ the header in the first file, and infer the types for the columns
	df_schema = df.schema

	##### SAVE JSON SCHEMA INTO S3 / BLOB STORAGE
	# save the schema to load from the streaming job, which we will load during the next job
	dbutils.fs.rm("/home/mwc/airline_schema.json", True)

	with open("/dbfs/home/mwc/airline_schema.json", "w") as f:



	lib <- .libPaths()[1]
	pkgs <- rownames(installed.packages(lib))

	install.packages(
	lib = lib,
	pkgs = as.data.frame(pkgs, stringsAsFactors=FALSE)$Package,
	type = 'source'
	)