Cosimo Lupo anthrotype

LLM Wiki v2

A pattern for building personal knowledge bases using LLMs. Extended with lessons from building agentmemory, a persistent memory engine for AI coding agents.

This builds on Andrej Karpathy's original LLM Wiki idea file. Everything in the original still applies. This document adds what we learned running the pattern in production: what breaks at scale, what's missing, and what separates a wiki that stays useful from one that rots.

What the original gets right

The core insight is correct: stop re-deriving, start compiling. RAG retrieves and forgets. A wiki accumulates and compounds. The three-layer architecture (raw sources, wiki, schema) works. The operations (ingest, query, lint) cover the basics. If you haven't read the original, start there.

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.

Index of FontTools extensions and utility libraries

afdko

https://github.com/adobe-type-tools/afdko/ Adobe Font Development Kit for OpenType

arabic-reshaper

https://github.com/mpcabd/python-arabic-reshaper/ Reconstruct Arabic sentences to be used in applications that don't support Arabic

Thoughts on fontTools.ufoLib

w.r.t. my experience with defcon and ufoLib. Some of the functionality I discuss (notifications, etc.) definitely shouldn't go into fontTools however we ought to make fontTools.ufoLib "compatible" with these extra features. I'm sure we can do simple and versatile.

IMO fontTools.ufoLib should basically be written from scratch (with copy-pasting here and there) since ufoLib/defcon have significant bloat and apparently we want to use lxml.

	#!/bin/bash
	#
	# Claude Code LSP Fix
	# ====================
	# Fixes the LSP plugin bug: https://github.com/anthropics/claude-code/issues/13952
	#
	# THE BUG:
	# Claude Code's LSP manager has an empty initialize() function that should
	# load and register LSP servers from plugins, but instead does nothing.
	# This causes "No LSP server available for file type" errors.

	from __future__ import print_function, division, absolute_import
	#from fontTools.misc.py23 import *

	class UnionFind(object):

	def __init__(self, items):
	self._sets = dict((x,set([x])) for x in items)
	self._map = dict((x,x) for x in items)

	def _union_pair(self, a, b):

	import collections
	import io
	import itertools
	import logging
	import os
	import sys
	from pathlib import Path
	from pprint import pprint
	from typing import Dict, List, Tuple

	import pathops
	from booleanOperations.booleanOperationManager import BooleanOperationManager
	import time

	import math

	print("pathops version:", pathops.__version__)

	f = CurrentFont()

	"""
	This compares the speed of experimental changes in ufoLib
	with the experimental ufoLib in fontTools. This requires the
	"Roboto-Regular.ufo" font to be located next to this script.
	"""

	import os
	import shutil
	import timeit
	import cProfile