drnic

import { Controller } from "@hotwired/stimulus"
import SignaturePad from 'signature_pad'

export default class extends Controller {
  static targets = ["canvas", "input"]
  connect() {
    this.signaturePad = new SignaturePad(this.canvasTarget)
    this.signaturePad.addEventListener("endStroke", this.endStroke)
    this.resizeCanvas()
    if (this.inputTarget.value) {

marckohlbrugge

🆕 Update: See more extensive repo here: https://github.com/marckohlbrugge/unofficial-37signals-coding-style-guide

The Unofficial 37signals/DHH Rails Style Guide

About This Document

This style guide was generated by Claude Code through deep analysis of the Fizzy codebase - 37signals' open-source project management tool.

Why Fizzy matters: While 37signals has long advocated for "vanilla Rails" and opinionated software design, their production codebases (Basecamp, HEY, etc.) have historically been closed source. Fizzy changes that. For the first time, developers can study a real 37signals/DHH-style Rails application - not just blog posts and conference talks, but actual production code with all its patterns, trade-offs, and deliberate omissions.

jnsahaj

# Create a new clone and branch for parallel development.
# Usage: ga <branch-name> [base-branch]
ga() {
  if [[ -z "$1" ]]; then
    echo "Usage: ga <branch-name> [base-branch]"
    return 1
  fi
  local branch="$1"
  local repo_name="$(basename "$PWD")"
  local repo_url="$(git remote get-url origin)"

karpathy

"""
The most atomic way to train and run inference for a GPT in pure, dependency-free Python.
This file is the complete algorithm.
Everything else is just efficiency.

@karpathy
"""

import os       # os.path.exists
import math     # math.log, math.exp

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.