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.

dabit3

See more of my writing here. Also check out Devin

In this post, I'll start from scratch and build up to OpenClaw's architecture step by step, showing how you could have invented it yourself from first principles, using nothing but a messaging API, an LLM, and the desire to make AI actually useful outside the chat window.

End goal: understand how persistent AI assistants work, so you can build your own (or become an OpenClaw power user).

First, let's establish the problem

When you use ChatGPT or Claude in a browser, there are several limitations:

dhh

# Create a new worktree and branch from within current git directory.
ga() {
  if [[ -z "$1" ]]; then
    echo "Usage: ga [branch name]"
    exit 1
  fi

  local branch="$1"
  local base="$(basename "$PWD")"
  local path="../${base}--${branch}"

martinwoodward

```mermaid
sequenceDiagram
    participant dotcom
    participant iframe
    participant viewscreen
    dotcom->>iframe: loads html w/ iframe url
    iframe->>viewscreen: request template
    viewscreen->>iframe: html & javascript
 iframe->>dotcom: iframe ready

garystafford

curl --silent --compressed \
'https://aws.amazon.com/api/dirs/items/search?item.directoryId=aws-products&sort_by=item.additionalFields.productCategory&sort_order=asc&size=500&item.locale=en_US' \
| jq -r '.items[].item | .additionalFields.productCategory + " | " + .additionalFields.productName' \
| sort

tobilg

{
  "us-east-1": {
    "city": "Ashburn",
    "state": "Virginia",
    "country": "United States",
    "countryCode": "US",
    "latitude": 38.9445,
    "longitude": -77.4558029,
    "region": "North America",
    "iataCode": "IAD"

pahud

#!/bin/bash
# video demo - https://www.youtube.com/watch?v=Y8TyE_DNds8

mkdir ~/.tmp && cd $_

# install aws-cli v2
curl "https://awscli.amazonaws.com/awscli-exe-linux-x86_64.zip" -o "awscliv2.zip" && \
  unzip awscliv2.zip && \
  sudo ./aws/install

singledigit

# You don't need this, but here it is anyways :)
openapi: "3.0.1"
info:
  title: "Translation API (Async)"
  version: "1.0.0"

paths:
  /:
    post:
      summary: Create new translation request

dabit3

GraphQL

GraphQL Schema

type Customer {
  id: ID!
  email: String!
}

singledigit

AWSTemplateFormatVersion: '2010-09-09'
Transform: AWS::Serverless-2016-10-31
Description: Event Bridge -> API Gateway

Resources:
  RestApiGateway:
    Type: AWS::Serverless::Api
    Properties:
      StageName: Prod