Christos Zisopoulos christos

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.

Beast Mode

Beast Mode is a custom chat mode for VS Code agent that adds an opinionated workflow to the agent, including use of a todo list, extensive internet research capabilities, planning, tool usage instructions and more. Designed to be used with 4.1, although it will work with any model.

Below you will find the Beast Mode prompt in various versions - starting with the most recent - 3.1

Installation Instructions

Go to the "agent" dropdown in VS Code chat sidebar and select "Configure Modes".
Select "Create new custom chat mode file"

Namespaces 101

Introduction

Ruby has recently merged namespaces as an experimental feature that is disabled by default, as of this writing.

This is a non-trivial development driven by @matz himself, and mainly implemented by @tagomoris, who just became a Ruby committer (🎉).

The feature has been cooking for a long time, with a first ticket opened a couple of years ago (#19744) and a revised one opened just last week (#21311).

Honest is looking for a Rails contractor (3-month, 40h/week commitment)

Honest is building the “customer service agent’s best friend”: contextual, real-time assistance and shortcuts for digital customer service teams.

We’re looking for an experienced Rails contractor to come help us improve our Rails app, initially with a 3-month, 40h/week commitment. We’ll appreciate an inclination for “foundation” work (enhancing performance, increasing test coverage, improving our CI/CD pipeline…) although there’s plenty to do in user-facing features too.

Honest is built as a pretty standard Rails 6 “beautiful monolith” hosted in Github and running on Heroku, powered by PostgreSQL and Redis. You’ll have the chance to acquire experience on newer stuff like Hotwire too.

You’ll be part of a small, fully-remote team coordinating mainly via Discord and Jira. You’ll work along with colleagues with plenty of experience in the different aspects of building digital products. We keep a healthy, respectful,

	#!/bin/sh

	# dfn: df normalized, or df not noisy, or df nicely, or df new
	#
	# Whatever you want to call it, this script makes df useful again.
	# It removes read-only disks for Time Machine backups, system volumes
	# that can't be modified, unbrowsable volumes, and other stuff you don't
	# really care about.
	#
	# Instead, you get the same drives that you see on your Mac desktop in an

	# I would really like to introspect complex authorization rules to understand why users can/cannot perform an action.
	# That helps with debugging and offers better error messages.

	# Sample:
	#
	# ```ruby
	# auth = Authorize.new(CanUpdate, owner, message_1)
	# pp auth.permitted? # => false
	# pp auth.explain
	# # => "(❌ IsAdmin) or (✅ IsPublisher and ❌ IsOwner)"

	require "json"

	struct = { "a" => 1, "b" => 2, "c" => [1, 2, 3], "d" => [{ "e" => 3 }, nil, false, true, [], {}] }
	source = JSON.dump(struct)

	tokens = []
	index = 0

	until source.empty?
	tokens <<

	# WAFs are a waste of time. False positives negatively impact actual users.
	# And for what? Protection against vulnerabilities that, if you still have
	# in this day in age, you're in more trouble than you think.
	# Still, to keep my logs clean, I like to run something like the follwoing:

	location ~ \.php$ { access_log off; return 444; }
	location ~ \.asp$ { access_log off; return 444; }
	location ~ \.aspx$ { access_log off; return 444; }
	location ~ /Admincenter/ { access_log off; return 444; }
	location /console/ { access_log off; return 444; }

	# spec/system/support/login_helpers.rb
	# See this blog post for setup guide: https://evilmartians.com/chronicles/system-of-a-test-setting-up-end-to-end-rails-testing

	module LoginHelpers
	def login_as(user)
	# Craft session cookie to make request authenticated (to pass even routing constraints)
	# Compilation of these:
	# - https://dev.to/nejremeslnici/migrating-selenium-system-tests-to-cuprite-42ah#faster-signin-in-tests
	# - https://turriate.com/articles/2011/feb/how-to-generate-signed-rails-session-cookie
	# - https://github.com/rails/rails/blob/43e29f0f5d54294ed61c31ddecdf76c2e1a474f7/actionpack/test/dispatch/cookies_test.rb#L350

	/*
	ERB template chunk from The Feed's display of emails:

	<section class="postings postings--feed-style" id="postings"
	data-controller="pagination" data-pagination-root-margin-value="40px">
	<%= render partial: "postings/snippet", collection: @page.records, as: :posting, cached: true %>

	<%= link_to(spinner_tag, url_for(page: @page.next_param),
	class: "pagination-link", data: { pagination_target: "nextPageLink", preload: @page.first? }) unless @page.last? %>
	</section>