karpathy

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.

hula-prose

{"lastUpload":"2021-07-19T14:02:44.063Z","extensionVersion":"v3.4.3"}

brenopolanski

1. Select the object(s) to export
2. Open the document properties window (Ctrl+Shift+D)
3. Select "Resize page to drawing or selection"
4. File > Save As Copy...
5. Select Optimized SVG as the format if you want to use it on the web

dAnjou

(Tested with KeePassXC on Fedora 25)

By default when using GNOME Keyring you have a keyring that is unlocked when you log in (usually called "Login"). You can make use of that by storing a KeePass database password in this keyring and using it to automatically unlock your KeePass database.

Store the KeePass database password in GNOME Keyring. You'll have to set a label and at least one attribute/value pair. The label is displayed in a GNOME keyring manager (e.g. Seahorse), the attribute/value pair should be a unique identifier because it's needed for the lookup. I suggest to use keepass as attribute and the database name as value (make sure it doesn't contain any spaces).

secret-tool store --label="KeePass <database_name>" keepass <database_name>

Then create a script to launch and immediately unlock your KeePass database.

zengabor

import unittest
unittest.TestLoader.testMethodPrefix = 'should'
from unittest import TestCase as Specification


class MyClass(Specification):
  
    def should_do_something():
        """This method with the 'should' prefix will be picked up by test runners"""