Brandon Summers evopix

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.

AI Agent Rule / Instruction / Context files / etc

Some notes on AI Agent Rule / Instruction / Context files / etc.

Adaptive Container Scaling

Problems with AWS's recommended scaling method

Based on a fixed percentage of memory and CPU reservation.
Scaling on both cpu and memory metrics can cause the metrics to fight with each other where it'll scale out then back in and repeat.
If you're scaling threshold is at 80% memory, current reservation utilization is at 79% and you deploy a container that requires 25% memory, It will fail to launch.
ASGs are not container aware.

Solution

Install linux subsystem Ubuntu 18.04

See Microsoft's WSL install guide for windows 10 for details.

Work-in-progress - command line only install:

Open Powershell as Administrator and run:
Enable-WindowsOptionalFeature -Online -FeatureName Microsoft-Windows-Subsystem-Linux
Open Windows App Store
- Search for "Ubuntu" and install Ubuntu 18.04

(work in progress - command line install steps):

Deploy updates to aws ecs services based on github commits. Posts deploy notifications to slack.

Assumptions:

Task definitions map 1:1 with services, and they have the same name.
Code is stored on github
You want to deploy the latest commit in a branch
Docker images are tagged with commit SHA
Docker images are stored on AWS ECR

	<template name="bool:default" value="/** * @var bool * @ORM\Column(type="boolean", options={"default": $default$}) */ private bool $$$name$ = $default$; $END$ public function is$method$(): bool { return $this->$name$; } public function set$method$(bool $$$name$): void { $this->$name$ = $$$name$; }" description="boolean column" toReformat="false" toShortenFQNames="true">
	<variable name="name" expression="" defaultValue="" alwaysStopAt="true" />
	<variable name="default" expression="enum("false", "true")" defaultValue="" alwaysStopAt="true" />
	<variable name="method" expression="capitalize(name)" defaultValue="" alwaysStopAt="false" />
	<context>
	<option name="PHP Class Member" value="true" />
	</context>
	</template>
	<template name="bool:nullable" value="/** * @var bool\|null * @ORM\Column(type="boolean", nullable=true) */ private ?bool $$$name$ = nul

	resource "aws_alb_listener_rule" "http_80" {
	count = local.lb ? 1 : 0
	listener_arn = data.aws_alb_listener.http_80.arn
	action {
	type = "forward"
	target_group_arn = aws_alb_target_group.http_80[0].arn
	}
	condition {
	host_header {
	values = [local.host_name]

	Mute these words in your settings here: https://twitter.com/settings/muted_keywords

	ActivityTweet
	generic_activity_highlights
	generic_activity_momentsbreaking
	RankedOrganicTweet
	suggest_activity
	suggest_activity_feed
	suggest_activity_highlights
	suggest_activity_tweet

	# use ImageMagick convert
	# the order is important. the density argument applies to input.pdf and resize and rotate to output.pdf
	convert -density 90 input.pdf -rotate 0.5 -attenuate 0.2 +noise Multiplicative -colorspace Gray output.pdf