stefjow

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.

Can We Get More Decentralised Than The Fediverse?

I guess that the [fediverse][fediverse] will be as decentralised as email: a bit, but not that much. Most people will be dependent on a few major hubs, some groups might have their own hubs (e.g. company email servers), personal instances will be pretty rare. This is in contrast to personal blogging, where every Bob can easily host their own (and they often do). I mean that's already implied by the name: fediverse is [a federated universe, not a distributed one][fed-v-dis].

Why does this matter? Well I like not being dependent on one entity, but I would like it much more if I was dependent on no entities at all. In other words, I like to publish my own personal blog and get all the goodies of a social network, without being dependent on other micro-blogging / social content platforms.

So in this writing, I'm going to:

❓ Contemplate on why the fediverse gets federated not distributed (spoilers: its push vs pull)
🧠 Ideate on how could we get a distri

Using Git Commit Message Templates to Write Better Commit Messages

The always enthusiastic and knowledgeable mr. @jasaltvik shared with our team an article on writing (good) Git commit messages: How to Write a Git Commit Message. This excellent article explains why good Git commit messages are important, and explains what constitutes a good commit message. I wholeheartedly agree with what @cbeams writes in his article. (Have you read it yet? If not, go read it now. I'll wait.) It's sensible stuff. So I decided to start following the

This guide now lives at https://github.com/joyeusenoelle/GuideToMastodon/ - please check there for updates, and feel free to submit a PR if you have suggestions or want to submit a translation!

	// Open the Chrome inspector, and select the topmost div containing the twitter thread.
	// You might want to open the first tweet in the thread, scroll down to load every item in the thread, then select the parent container

	var foo = document.createElement("div");
	var str = ""
	$($0).find(".tweet").each(function(i, t) {
	var tweet = $(t);
	var turl = "https://twitter.com" + tweet.data("permalink-path")
	var tdate = tweet.find(".tweet-timestamp").attr("title")
	var tcontent = tweet.find(".tweet-text").html()

	#!/bin/bash
	#
	# Requires:
	# - gdal_sieve.py
	# - ogr2ogr (GDAL)
	# - topojson (node.js)

	# Grab the relative directory for source file.
	SRC_DIR=`dirname $0`

stefjow

LLM Wiki

The core idea

Can We Get More Decentralised Than The Fediverse?

Using Git Commit Message Templates to Write Better Commit Messages

Table of Contents