Louie de la Rosa luisdelarosa

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.

	# Add this snippet to the top of your playbook. We'll refer to this as "your-playbook.yml"
	# It will install python2 if missing (but checks first so no expensive repeated apt updates)
	# [email protected]
	# Assumes that you are not running as root
	# Run via: "ansible-playbook --ask-become-pass your-playbook.yml"
	# v1.1 by [email protected] / @louielouie

	- hosts: all
	gather_facts: False

	// This sqlite3_exec limits the database to 50 DB pages (about 1.5Kb each)
	// Huge memory win on iPhone.
	if(sqlite3_exec(myDatabsae, "PRAGMA CACHE_SIZE=50", NULL, NULL, NULL) != SQLITE_OK)
	NSLog(@"Couldn't set cache size: %s", sqlite3_errmsg([self.database sql]));

	@implementation NSImage (RSShading)

	- (NSImage *) tintedImage
	{
	NSImage* newImage = [[[NSImage alloc] initWithSize:[self size]] autorelease];
	NSRect imageRect = {NSZeroPoint, [self size]};
	NSColor *color = [NSColor blackColor];

	[newImage lockFocus];
	[[color colorWithAlphaComponent: 0.25] set];

	// Use Gists to store code you would like to remember later on
	console.log(window); // log the "window" object to the console