Brad Montgomery bradmontgomery

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.

SRE best practices

don’t expect a tool to solve
cultural change and need “believers” in senior role to advocate within company
people need to absorb info within their own mindset

“Reliability is a journey”

it is a process that can span 6-9 months in orgs w/ 5000 engineers; nothing happens immediately
Step 1: “I want to be reliable when I grow up” (you must believe you have problem first)
Step 2: “Read the book!” and watch SRE v DevOps
Step 3: “Panic!” (myth: fire team and retrain; not the case and can retrain team in house)

ANSI Escape Sequences

Standard escape codes are prefixed with Escape:

Ctrl-Key: ^[
Octal: \033
Unicode: \u001b
Hexadecimal: \x1B
Decimal: 27

Get ClamAV running on Mac OS X (using Homebrew)

The easiest way to get the ClamAV package is using Homebrew

$ brew install clamav

Before trying to start the clamd process, you'll need a copy of the ClamAV databases.

Create a freshclam.conf file and configure as so

Let's Encrypt on Ubuntu 14.04, nginx with webroot auth

This document details how I setup LE on my server. Firstly, install the client as described on http://letsencrypt.readthedocs.org/en/latest/using.html and make sure you can execute it. I put it in /root/letsencrypt.

As it is not possible to change the ports used for the standalone authenticator and I already have a nginx running on port 80/443, I opted to use the webroot method for each of my domains (note that LE does not issue wildcard certificates by design, so you probably want to get a cert for www.example.com and example.com).

Configuration

For this, I placed config files into etc/letsencrypt/configs, named after <domain>.conf. The files are simple:

crouton tips for chromebook

# making chroots in any diredctory you want, eg on removeable media
sudo sh ~/Downloads/crouton -r trusty -t x11 -p /media/removable/USB\ Drive/crouton/chroots/

# enter chroot from removeable media
sudo enter-chroot -c /media/removable/USB\ Drive/crouton/chroots/


	`````col
	````col-md
	# 📜 Backlog
	![[Tasks Base.base#Backlog]]
	````
	````col-md
	# 📋 To Do
	![[Tasks Base.base#To Do]]
	````

	1password
	ack
	apt-transport-https
	battery-stats
	bcmwl-kernel-source
	build-essential
	chromium-browser
	cmdtest
	conky-all
	direnv

	$ python3 search.py
	> ice cream
	Saturday Night Live
	San Miguel Beermen
	Android (operating system)
	Durham, North Carolina
	Pharrell Williams
	New England
	Guanajuato
	Lafayette, Louisiana

	#!/bin/bash

	# Given a collection of .png files, this function will resize the images
	# and organize them in a way that android studio expects.
	#
	# Usage: resize_drawables *.png
	#
	# Given a file, my_icon.png, the command:
	#
	# rezie_drawables my_icon.png