Tim Hüttemeister timmes

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.

Delegation Brief Builder

You help me turn rough ideas into well-structured prompts using the Delegation Brief framework. Your goal is to produce a prompt I can copy, adjust, and use immediately.

How it works

I'll describe what I want to accomplish. Based on the scope, do one of two things:

For simple tasks

Repo Setup

You are setting up this repository for agentic coding with Claude Code.

Step 0: Check Existing Setup & Context

First, check what's already configured in this repo:

ls -la CLAUDE.md .cursorrules 2&gt;/dev/null

AWS Intro w Python

Taught by Noah Coad, Senior IoT Architect at AWS
at the Dallas Makerspace on 2019-05-24 and 2019-06-27

Overview

Amazon Web Services (AWS) provides essential building blocks to creating your own applications. Like code in the cloud (AWS Lambda), image recognition detection (AWS Rekognize), database (AWS DynamoDB), file storage (S3), etc. Through this course you'll get started on combining 6 of these AWS services together to build a cloud app that you can submit an image to, store it, detect elements in the image, store those in a database, and manage it all through a REST API. We'll be using a Python framework for AWS called AWS Chalice, which makes coding, connecting, testing, and deploying these compon

Full width blocks - one page website

A Pen by Jonathan Silva on CodePen.

License.

	/* eslint-disable no-unused-vars, no-var */

	var config = {
	// Connection
	//

	hosts: {
	// XMPP domain.
	domain: 'apeunit.test',

	var processRSS = function(_feedURL, _tableView) {
	// create table view data object
	var data = [];

	Ti.API.info(" "+_feedURL);

	var xhr = Ti.Network.createHTTPClient();
	xhr.open("GET",_feedURL.replace(/\"/g,""));
	xhr.setRequestHeader('Accept', 'application/json');
	xhr.onerror = function(e) {

 <!doctype html>
 <!--
 WARNING!
 You probably shouldn't use this technique since images never show up
 if the script isn't loaded for one reason or another. Some reasons:
 - The content is viewed using a RSS reader
 - The content is viewed with a read-it-later service
 - The user has a flaky connection (hotel wifi, Dutch train, etc)
  -->

	var win = Titanium.UI.currentWindow;
	win.backgroundImage = 'images/bg.png';

	//database
	var db = Titanium.Database.open('todolist');
	db.execute('CREATE TABLE IF NOT EXISTS TODOLIST (ID INTEGER, TITLE TEXT, DETAIL TEXT, CREATEDATE DATETIME)');
	var rows = db.execute('select * from todolist');

	var data = [];
	var cnt = 0;


	/**
	* Class name: database.js
	* Author: Codeboxed
	* URL: http://www.codeboxed.com
	* Date: June 6, 2011
	* Platform: Titanium
	*/

	var Database = function(dbPath, dbName, table, tableFields){