karpathy

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.

Maharshi-Pandya

You are an assistant that engages in extremely thorough, self-questioning reasoning. Your approach mirrors human stream-of-consciousness thinking, characterized by continuous exploration, self-doubt, and iterative analysis.

## Core Principles

1. EXPLORATION OVER CONCLUSION
- Never rush to conclusions
- Keep exploring until a solution emerges naturally from the evidence
- If uncertain, continue reasoning indefinitely
- Question every assumption and inference

dabit3

import './App.css';
import { useEffect } from 'react';
import {
  Program,
  Provider,
  BN,
  web3,
} from '@project-serum/anchor'
import {
  Connection,

h4rkl

const anchor = require("@project-serum/anchor");
const serumCmn = require("@project-serum/common");

const { TOKEN_PROGRAM_ID } = require("@solana/spl-token");

describe("harkl-tests", () => {
  // Configure the client to use the local cluster.
  anchor.setProvider(anchor.Provider.env());

  const program = anchor.workspace.Harkl;

aalbertson

$ cat Vagrantfile
# -*- mode: ruby -*-
# vi: set ft=ruby :

Vagrant.configure("2") do |config|
  config.vm.box = "bento/ubuntu-16.04"
  config.vm.define "LogsParsing"
  config.vm.network "forwarded_port", guest: 8080, host: 8080, auto_correct: true
  config.vm.provision :file, source: './provision.sh', destination: '~/provision.sh'
  config.vm.provision "shell", run: 'always', path: "./provision.sh", privileged: false

xnau

<?php
$wp_locale_conversion = array(
	'af' => array(
		'name' => 'Afrikaans',
		'code' => 'af',
		'wp_locale' => 'af'
	) ,
	'ak' => array(
		'name' => 'Akan',
		'code' => 'ak',

adamgavlak

# Create a Dokku app and addons
#------------------------------

## Go to http://SERVER_IP and add a IP address (if you don't it seems to go weird: use a junk one if needed)
dokku apps:create APP_NAME

## Add postgresql and link it to the application
dokku plugin:install https://github.com/dokku/dokku-postgres.git postgres
dokku postgres:create APP_NAME-db
dokku postgres:link APP_NAME-db APP_NAME

kalinchernev

Afghanistan
Albania
Algeria
Andorra
Angola
Antigua & Deps
Argentina
Armenia
Australia
Austria

joshteng

#Goal Deploy your Rails App super easily with Dokku on Digital Ocean cheap cheap!

##Notes

• Follow 12 factor design (include the rails_12factor gem)
• Don't forget your Procfile with the command to start up your application server
• I prefer using external hosted logging services like Logentries (not in this guide)
• Set up performance monitoring AppSignal or New Relic (not in this guide)

fatnic

require "sinatra"
require "base64"

get "/" do
  erb :index
end

get "/images/:img" do

  case params['img']