Skip to content

Instantly share code, notes, and snippets.

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.

@timotgl
timotgl / uninstall-logitech-ghub.sh
Created April 4, 2022 12:15
How to fully uninstall Logitech G HUB on macOS via terminal/command line
# How to fully uninstall Logitech G HUB on macOS via terminal/command line
# Tested on macOS version 12.3.1 (21E258) Monterey in April 2022
# with Logitech G HUB version 2022.3.242300 (released on 2022-03-22) installed.
# 1. Make sure "Logitech G HUB" itself is not running. If it is, quit it.
# 2. Open "Activity Monitor" and force-quit all processes named "lghub*".
# 3. Delete system-wide files
sudo rm -rf /Applications/lghub.app
#!/bin/bash
# Like the 'aws' cli except caches results
# http://redsymbol.net/articles/unofficial-bash-strict-mode/
set -euo pipefail
IFS=$'\n\t'
#set -x
@paxan
paxan / get-daemontools-heroku-16.md
Created March 9, 2017 18:15
How to get daemontools onto a dyno running heroku-16 stack
curl -sO https://mirrors.kernel.org/ubuntu/pool/universe/d/daemontools/daemontools_0.76-6ubuntu1_amd64.deb
dpkg-deb -x daemontools_0.76-6ubuntu1_amd64.deb daemontools
@yossorion
yossorion / what-i-wish-id-known-about-equity-before-joining-a-unicorn.md
Last active September 8, 2026 14:07
What I Wish I'd Known About Equity Before Joining A Unicorn

What I Wish I'd Known About Equity Before Joining A Unicorn

Disclaimer: This piece is written anonymously. The names of a few particular companies are mentioned, but as common examples only.

This is a short write-up on things that I wish I'd known and considered before joining a private company (aka startup, aka unicorn in some cases). I'm not trying to make the case that you should never join a private company, but the power imbalance between founder and employee is extreme, and that potential candidates would

@paxan
paxan / ae.clj
Last active November 25, 2016 21:31
AEAD decryption process using AWS KMS & AES/GCM/NoPadding cipher
;; CAVEAT: Ensure your JDK/JRE is configured with Java Cryptography
;; Extension (JCE) Unlimited Strength Jurisdiction Policy Files.
;; Visit http://www.oracle.com/technetwork/java/javase/downloads/index.html
;; and look for "JCE".
;; Given a ciphertext (produced by AES/GCM/NoPadding cipher), a nonce,
;; and a KMS-encrypted content encrypting key (cek), here is the
;; process for producing the plaintext:
(defn b64dec [x] (javax.xml.bind.DatatypeConverter/parseBase64Binary x))
@paxan
paxan / tornado_force_https.py
Last active August 15, 2023 01:24
A hack to make Tornado web server redirect http requests to https for Heroku or similar reverse-proxied deployments
from __future__ import absolute_import, print_function
import os
import tornado.httpserver
import tornado.ioloop
import tornado.web
def create_server(*args, **kwargs):
'''
@mingfang
mingfang / convert id_rsa to pem
Last active December 12, 2024 12:13
Convert id_rsa to pem file
openssl rsa -in ~/.ssh/id_rsa -outform pem > id_rsa.pem
chmod 600 id_rsa.pem
@staltz
staltz / introrx.md
Last active October 28, 2026 13:21
The introduction to Reactive Programming you've been missing
@igrigorik
igrigorik / github.bash
Last active September 12, 2025 09:24
Open GitHub URL for current directory/repo...
alias gh="open \`git remote -v | grep git@github.com | grep fetch | head -1 | cut -f2 | cut -d' ' -f1 | sed -e's/:/\//' -e 's/git@/http:\/\//'\`"