Florian Reichelt FReichelt

Semantic Commit Messages

See how a minor change to your commit message style can make you a better programmer.

Format: <type>(<scope>): <subject>

<scope> is optional

Example

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.

	#!/bin/bash
	# Launches socat+npiperelay to relay the gpg-agent socket file for use in WSL
	# See https://justyn.io/blog/using-a-yubikey-for-gpg-in-windows-10-wsl-windows-subsystem-for-linux/ for details

	GPGDIR="${HOME}/.gnupg"
	USERNAME=Vincent
	# I use the same username for wsl and windows, but feel free to modify the paths below if that isn't the case
	WIN_GPGDIR="C:/Users/${USERNAME}/AppData/Roaming/gnupg"
	NPIPERELAY="${HOME}/npiperelay.exe"

	// This is based on original code from http://stackoverflow.com/a/22649803
	// with special credit to error454's Python adaptation: https://gist.github.com/error454/6b94c46d1f7512ffe5ee

	function EnhanceColor(normalized) {
	if (normalized > 0.04045) {
	return Math.pow( (normalized + 0.055) / (1.0 + 0.055), 2.4);
	}
	else { return normalized / 12.92; }

	}