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.

ajkerrigan

#!/bin/bash
#
# Mush mosh and SSM sessions together.
#
# Run mosh-server on an EC2 instance as part of an AWS Session Manager SSH
# session, and feed the output into mosh-client locally. This is helpful
# if you have an EC2 instance which:
#
# * Is publicly addressable
# * Has a predictable Name tag but a shifting IP

gfrlv

FROM alpine:3.8 as build

ENV CXXFLAGS -std=c++98

RUN apk add --no-cache \
        gcc g++ musl-dev make libc6-compat cmake \
        unixodbc-dev zlib-dev openssl-dev ncurses-dev bison \
        wget

lisawolderiksen

Using Git Commit Message Templates to Write Better Commit Messages

The always enthusiastic and knowledgeable mr. @jasaltvik shared with our team an article on writing (good) Git commit messages: How to Write a Git Commit Message. This excellent article explains why good Git commit messages are important, and explains what constitutes a good commit message. I wholeheartedly agree with what @cbeams writes in his article. (Have you read it yet? If not, go read it now. I'll wait.) It's sensible stuff. So I decided to start following the

egmontkob

Hyperlinks (a.k.a. HTML-like anchors) in terminal emulators

[ Update 2025-03-24: Commenting is disabled permanently. Previous comments are archived at web.archive.org. ]

---

Most of the terminal emulators auto-detect when a URL appears onscreen and allow to conveniently open them (e.g. via Ctrl+click or Cmd+click, or the right click menu).

It was, however, not possible until now for arbitrary text to point to URLs, just as on webpages.

JayH5

#!/usr/bin/env python
import argparse
import base64
import json
import os


def main():
    parser = argparse.ArgumentParser(
        description="Dump all certificates out of Traefik's acme.json file")

guillemcanal

Build PHP5 iconv extension from source for Alpine

If you plan to use iconv() to transliterate string in you PHP project, please, take this:

FROM alpine:3.4

RUN apk add --update php5-cli wget build-base php5-dev autoconf re2c libtool \

sigma

{ pkgs ? import <nixpkgs> {} }:

with pkgs;
let

  su_exec = pkgs.stdenv.mkDerivation {
    name = "su-exec-0.2";
    src = fetchurl {
      url = https://github.com/ncopa/su-exec/archive/v0.2.tar.gz;
      sha256 = "09ayhm4w7ahvwk6wpjimvgv8lx89qx31znkywqmypkp6rpccnjpc";