Justin Tanner justintanner

Running Gas City across multiple cities and providers

Now available as a repo: https://github.com/thinkjones/gascity-cookbook

A worked example of a multi-city, multi-provider Gas City setup: one config monorepo that runs several "cities" (agent deployments), each on a different model provider, sharing one reusable pack of config.

The trick that makes it maintainable is a tier abstraction: your agents and workflows only ever refer to abstract tiers — fast, solid, deep — and each city binds those

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.

How to deploy a Rails 7.1 app with Postgres and Kamal on a single server

I think you have looked at the tutorial from Mr. Heinemeier Hansson at least once or twice and have a similar setup.

rails new kamal_pg --css tailwind --skip-test --database=postgresql

cd kamal_pg

This demonstrates the implementation of full text search for documents in Indexed DB.

Word-breaking and stemming is used to create a list of terms for each document.
Document records are annotated with the list of terms when added to the database.
A multi-entry index on the list of terms is populated.
A query is similarly processed into a list of terms.
A join over the terms is implemented using multiple cursors on the index.

The necessity of annotating records with the word list to populate the index is a limitation of the current Indexed DB API. A feature request to support custom

	# Dont forget to set the env variable "certdomain", and either fill in your email below or use an env variable for that too.
	# Also note that this config is using the LetsEncrypt staging server, remove the flag when ready!

	Resources:
	sslSecurityGroupIngress:
	Type: AWS::EC2::SecurityGroupIngress
	Properties:
	GroupId: {"Fn::GetAtt" : ["AWSEBSecurityGroup", "GroupId"]}
	IpProtocol: tcp
	ToPort: 443

	<?xml version="1.1" encoding="UTF-8"?>
	<!DOCTYPE keyboard SYSTEM "file://localhost/System/Library/DTDs/KeyboardLayout.dtd">
	<!--

	Data generated Wed Mar 23 01:35:41 2022

	Generated by kluchrtoxml_64 build 203

	-->
	<!--Last edited by Ukelele version 351 on 2022-03-23 at 01:45 (GMT+8)-->

	const crypto = require('crypto');

	const PASSWORD = "098f6bcd4621d373cade4e832627b4f6"
	, MESSAGE = 'test';

	function InvalidSignatureError() {
	Error.captureStackTrace(this, this.constructor);
	}

	function encipher(message, password, callback) {

	# in features/support/env.rb
	require 'selenium/webdriver'

	# we need a firefox extension to start intercepting javascript errors before the page
	# scripts load
	Capybara.register_driver :selenium do \|app\|
	profile = Selenium::WebDriver::Firefox::Profile.new
	# see https://github.com/mguillem/JSErrorCollector
	profile.add_extension File.join(Rails.root, "features/support/extensions/JSErrorCollector.xpi")
	Capybara::Selenium::Driver.new app, :profile => profile

	{
	/* ~/Library/KeyBindings/DefaultKeyBinding.dict

	See https://gist.github.com/zsimic/1367779, prefix cheatsheet:
	Char Short Key
	$ S- Shift ⇧
	^ C- CTRL ⌃
	~ O- Option ⌥ (⊞ Win key position on PC)
	@ M- Command ⌘ (Alt position on PC)
	# Numeric Keypad