Nick Benes nickbenes

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.

Migrations and testing for PostgreSQL using node.js and Travis-CI

We're looking to add a column to the table user called admin (boolean) in an existing database.

Project layout

migrations

Creating Whispernote's GitHub Project Page

Jekyll, GitHub's static site generator, and Bootstrap were used to create Whispernote's GitHub Project Page. This document shows how.

Project Goals

I wanted a stylish "brand" page for my project, and I wanted to reuse the README file I had created without a lot of copy/pasting. GitHub's Jekyll allowed me to essentially embed the file as a partial. What I got was really an enhanced README file (thanks to some jQuery and Bootstrap), which is exactly what I wanted.

	var createThumb = function(fileObj, readStream, writeStream) {
	// Transform the image into a 10x10px thumbnail
	gm(readStream, fileObj.name()).resize('10', '10').stream().pipe(writeStream);
	};
	var createWeb = function(fileObj, readStream, writeStream) {
	// Transform the image into a 10x10px thumbnail
	gm(readStream, fileObj.name()).resize('320', '180').stream().pipe(writeStream);
	};
	var convertToPng = function(fileObj) {
	return {