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.

uchilaka

var MockPrompt = {
  install: function() {
    MockPrompt.message = undefined;
    MockPrompt.value = undefined;
    MockPrompt.returnValue = undefined;
    MockPrompt.originalPrompt = window.prompt;

    window.prompt = function(message, value) {
      MockPrompt.message = message;
      MockPrompt.value = value;

shwei

'use strict';

const functions = require('firebase-functions');

const fastify = require('./fastify');
exports.greetFromFastify = functions.https.onRequest(fastify);

uchilaka

module.exports = {
  parser: "babel-eslint",
  parserOptions: {
    ecmaVersion: 6,
    sourceType: 'module',
    ecmaFeatures: {
      modules: true,
      //experimentalObjectRestSpread: true
    }
  },

Chocksy

# FOR BUSY JOBS
# take the process_id from the /busy page in sidekiq and kill the longest running one.
workers = Sidekiq::Workers.new
long_process_id = 'integration.3:4:71111aaa111' # Eg: 'integration.3:4:71d1d7f4ef5a'
workers.each do |process_id, thread_id, work|
  process = Sidekiq::Process.new('identity' => process_id)
  process.stop! if process_id == long_process_id
end

# FOR SCHEDULED JOBS

ncochard

When you create a npm package, remember it might be used in a browser or a server, or even a command line utility… For each package you create, please pay attention at what it will be used for:

1. Is it going to be used as a dependency to a nodejs application that is not bundled? (e.g. command line utilities)
2. Is it going to be used as a dependency to a nodejs application that is bundled? (e.g. AWS Lambdas)
3. Is it going to be used as a dependency to a browser application (always bundled)?.

• In cases 2) and 3) you want to allow for tree shaking.
• In cases 1) and 2) you want to benefit from the "ES6"/"ES next" features supported natively by nodejs.
• In case 3) you also want to benefit from the native support of "ES6" from your browser.

dillansimmons

// JS for grabbing utm params and parsing into url
var getRefQueryParam = function() {
var temp = {};
document.location.search.replace(/\??(?:([^=]+)=([^&]*)&?)/g, function() {
var decode = function(s) {
return decodeURIComponent(s.split("+").join(" "));
};
temp[decode(arguments[1])] = decode(arguments[2]);
});
return temp;

rambabusaravanan

# Linux
# add the following to "~/.gitconfig" file

[merge]
    tool = intellij
[mergetool "intellij"]
    cmd = /usr/local/bin/idea merge $(cd $(dirname "$LOCAL") && pwd)/$(basename "$LOCAL") $(cd $(dirname "$REMOTE") && pwd)/$(basename "$REMOTE") $(cd $(dirname "$BASE") && pwd)/$(basename "$BASE") $(cd $(dirname "$MERGED") && pwd)/$(basename "$MERGED")
    trustExitCode = true

[diff]

anubhavshrimal

[
{
"name": "Afghanistan",
"dial_code": "+93",
"code": "AF"
},
{
"name": "Aland Islands",
"dial_code": "+358",
"code": "AX"

tonysneed

- Open Automator 
- File -> New -> Service 
- Change "Service Receives" to "files or folders" in "Finder" 
- Add a "Run Shell Script" action 
- Change "Pass input" to "as arguments" 
- Paste the following in the shell script box: open -n -b "com.microsoft.VSCode" --args "$*" 
- Save it as something like "Open in Visual Studio Code"