Abraham abranhe

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.

Pure ESM package

The package that linked you here is now pure ESM. It cannot be require()'d from CommonJS.

This means you have the following choices:

Use ESM yourself. (preferred)
Use import foo from 'foo' instead of const foo = require('foo') to import the package. You also need to put "type": "module" in your package.json and more. Follow the below guide.
If the package is used in an async context, you could use await import(…) from CommonJS instead of require(…).
Stay on the existing version of the package until you can move to ESM.

Convert Java Executable to Linux Executable

Introduction

Java program is universal and applicable to most operating systems, but if someone intends to create a native executable binary file for certain operation system, is that possible?

The answer is yes. Here is the instruction to create an executable program for Linux with .run as its extension name.

Create a text file use your favorite text editor (In this case it is gedit) and type these code below:

ES6 Setup

npm init -y
npm i --save-dev nodemon
npm add babel-preset-env babel-cli

Create a .babelrc config in your project root. Insert the following

{
 "presets": ["env"]

Handling Multiple Github Accounts on MacOS

The only way I've succeeded so far is to employ SSH.

Assuming you are new to this like me, first I'd like to share with you that your Mac has a SSH config file in a .ssh directory. The config file is where you draw relations of your SSH keys to each GitHub (or Bitbucket) account, and all your SSH keys generated are saved into .ssh directory by default. You can navigate to it by running cd ~/.ssh within your terminal, open the config file with any editor, and it should look something like this:

Host *
 AddKeysToAgent yes

> UseKeyChain yes

Semantic Commit Messages

See how a minor change to your commit message style can make you a better programmer.

Format: <type>(<scope>): <subject>

<scope> is optional

	(async () => {
	const FILTER = ""; // "" = all, or "arrow", "home", etc.
	const MAX_ICONS = Infinity;
	const ICON_SIZE = 96;
	const OPEN_LIBRARY_MENU = true;
	const DEFAULT_STATUS = "unpublished";

	const uid = () => crypto.randomUUID().replace(/-/g, "").slice(0, 21);
	const rand = () => Math.floor(Math.random() * 2 ** 31);
	const now = Date.now();

	import {useCallback, useRef, useEffect} from 'react'
	import {useField, useFormikContext} from 'formik'
	import update from 'immutability-helper'

	const useFieldArray = props => {
	const [field, meta] = useField(props)
	const fieldArray = useRef(field.value)
	const {setFieldValue} = useFormikContext()

	useEffect(() => {

	import java.lang.Math;

	public class Main {

	public static double inverseSqrt5 = 1 / Math.sqrt(5);
	public static double phi = (1 + Math.sqrt(5)) / 2;

	public static int fibonacci(int n) {
	return (int)Math.floor(Math.pow(phi, n) * inverseSqrt5 + 0.5);
	}

	// EmailInput wraps an HTML `input` and adds some app-specific styling.
	const EmailInput = React.forwardRef((props, ref) => (
	<input ref={ref} {...props} type="email" className="AppEmailInput" />
	));

	class App extends Component {
	emailRef = React.createRef();

	render() {
	return (