Tony R Quilkey trq

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.

Self-Improving Claude Code: A Bootstrap Seed

The Hypothesis

A single prompt (~1400 tokens), placed in a project's .claude/CLAUDE.md, can bootstrap a Claude Code instance into a self-improving system — one that captures learnings, extracts patterns, evolves its own configuration, and gets meaningfully better at helping its user with each session.

No pre-built infrastructure required. No user-level config. No hooks, skills, templates, or elaborate folder hierarchies. Just a seed and the affordances Claude Code already provides.

Background

Using Xdebug in Docker

Getting setup and running with Xdebug in a docker container these days is now fairly simple and is composed of two main steps:

Ensure you've got XDebug installed and enabled in your PHP docker image, for example:

# Installing Xdebug with PHP 7.3 images:
RUN pecl install xdebug \
  && docker-php-ext-enable xdebug

	version: '3'
	services:
	# FRONT
	chronograf:
	# Full tag list: https://hub.docker.com/r/library/chronograf/tags/
	image: chronograf
	deploy:
	replicas: 1
	placement:
	constraints:

	<?php namespace App\Finite;

	/**
	* The FiniteAuditTrail Trait.
	* This plugin, for the Finite package (see https://github.com/yohang/Finite), adds support for keeping an audit trail for any state machine.
	* Having an audit trail gives you a complete history of the state changes in your stateful model.
	* Prerequisites:
	* 1. Install Finite package (https://github.com/yohang/Finite#readme)
	* 2. Use FiniteStateMachine in your model (https://gist.github.com/tortuetorche/6365575)
	* Usage: in your Stateful Class use this trait after FiniteStateMachine trait, like this "use FiniteAuditTrailTrait;".

	#!/bin/bash

	current=$PWD

	cd $HOME

	sudo brew install zsh
	chsh -s /bin/zsh $USER

	brew tap thoughtbot/formulae

	#!/bin/bash

	current=$PWD

	cd $HOME

	sudo apt-get install zsh
	chsh -s /bin/zsh $USER

	wget https://thoughtbot.github.io/rcm/debs/rcm_1.3.0-1_all.deb

	#!/bin/sh

	#
	# Forwards SSH agent into a Docker container running in the active
	# docker-machine
	#

	PROGNAME=$(basename $0)

	NAME=$(docker-machine active)

	git remote prune origin
	git branch -r --merged master \| egrep -iv '(master\|develop)' \| sed 's/origin\///g' \| xargs -n 1 git push --delete origin

	#!/bin/bash
	# Coding Standards fixer from http://cs.sensiolabs.org/ pre-commit hook for git
	#
	# Based on https://github.com/s0enke/git-hooks
	#
	# @author Soenke Ruempler <soenke@ruempler.eu>
	# @author Sebastian Kaspari <s.kaspari@googlemail.com>
	#
	# see the README