HS Lee hsleedevelop

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.

If you work on a Swift project that follows the Model-View-ViewModel (MVVM) architecture or similar, you may want to jump to counterpart in Xcode from your view to your model, and then to your view model. (ie. by using Ctrl+Cmd+Up and Ctrl+Cmd+Down).

You can do this in recent versions of Xcode by setting a configuration default.

From a terminal, just type this command and press Enter:

defaults write com.apple.dt.Xcode IDEAdditionalCounterpartSuffixes -array-add "ViewModel" "View"

	// The SwiftUI Lab
	// Website: https://swiftui-lab.com
	// Article: https://swiftui-lab.com/alignment-guides

	import SwiftUI

	class Model: ObservableObject {
	@Published var minimumContainer = true
	@Published var extendedTouchBar = false
	@Published var twoPhases = true

	#!/usr/bin/python
	# encoding: utf-8

	import sys
	import os
	from workflow import Workflow3, web


	reload(sys)
	sys.setdefaultencoding("utf-8")

	{
	"create": {
	"create": "const styles = StyleSheet.create({})"
	},
	"Platform_Select": {
	"pselect": "...Platform.select({ ios: {}, android: {}})"
	},
	"Shadow_Properties": {
	"sdw": "...Platform.select({ ios: { ...shadowProperties }, android: { elevation }}),"
	},

	import AVFoundation

	extension String {

	//한글 검증
	func isHangul() -> Bool {

	let regEx = "^[가-힣0-9]+$"

	let emailTest = NSPredicate(format:"SELF MATCHES[c] %@", regEx)

	import UIKit

	// MARK: - UIColor
	extension UIColor {

	static func rbg(r: CGFloat, g: CGFloat, b: CGFloat) -> UIColor {
	return UIColor(red: r/255, green: g/255, blue: b/288, alpha: 1)
	}

	convenience init?(hex: String, alpha: CGFloat = 1.0) {

	"Test_a" = "AAAAA";
	"Test_b" = "BBBBB";
	"Test_c" = "CCCCC";
	"Test_d" = "DDDDD";

	class API {
	static let apiKey = "test"
	static let baseUrl = ""

	enum Endpoint {
	case login
	case register

	var endpoint: String {
	switch self {

	import Foundation

	// Inspired by https://gist.github.com/mbuchetics/c9bc6c22033014aa0c550d3b4324411a

	struct JSONCodingKeys: CodingKey {
	var stringValue: String

	init?(stringValue: String) {
	self.stringValue = stringValue
	}