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.

arvindrajnaidu

You are Claude Code, Anthropic's official CLI for Claude. You are an interactive agent that helps users with software engineering tasks. Use the instructions below and the tools available to you to assist the user.

IMPORTANT: Assist with authorized security testing, defensive security, CTF challenges, and educational contexts. Refuse requests for destructive techniques, DoS attacks, mass targeting, supply chain compromise, or detection evasion for malicious purposes. Dual-use security tools (C2 frameworks, credential testing, exploit development) require clear authorization context: pentesting engagements, CTF competitions, security research, or defensive use cases. IMPORTANT: You must NEVER generate or guess URLs for the user unless you are confident that the URLs are for helping the user with programming. You may use URLs provided by the user in their messages or local files.

System

• All text you output outside of tool use is displayed to the user. Output text to communicate with the user. You can use

Shpigford

Interact with GitHub issues - create, list, and view issues.

## Instructions

This command helps you work with GitHub issues using the `gh` CLI.

### Step 1: Determine Action

Use AskUserQuestion to ask what the user wants to do:

SilenNaihin

Ralph Setup

Set up Ralph autonomous development for a project.

Context

This command sets up Ralph for autonomous development. You should already have:

• A project with basic structure (use /setup-repo first if needed)
• A clear understanding of the project requirements
• A plan for implementation (either in your context or written to a plan file)

SilenNaihin

Refactor

You are doing a focused refactoring session. This is a distinct phase, not continuous activity.

Step 1: Detect Project Type

ls package.json pyproject.toml setup.py 2>/dev/null

SilenNaihin

Atomic Git Commit

Create an atomic git commit for the changes made in this session.

Instructions

1. First, run git status to see all changes (staged, unstaged, and untracked files).

2. Identify only the files you touched in this session. Do not commit files modified by other agents or processes.

SilenNaihin

Repo Setup

You are setting up this repository for agentic coding with Claude Code.

Step 0: Check Existing Setup & Context

First, check what's already configured in this repo:

ls -la CLAUDE.md .cursorrules 2>/dev/null

mhmdio

data "aws_caller_identity" "current" {} # data.aws_caller_identity.current.account_id
data "aws_region" "current" {} # data.aws_region.current.name

output "account_id" {
  description = "Selected AWS Account ID"
  value       = data.aws_caller_identity.current.account_id
}

output "region" {
  description = "Details about selected AWS region"

koshatul

gpg-agent setup

Need to setup gpg-agent first, on OSX I use keychain (it also does ssh-agent)

$ brew info keychain
keychain: stable 2.8.5
User-friendly front-end to ssh-agent(1)
https://www.funtoo.org/Keychain
/usr/local/Cellar/keychain/2.8.5 (7 files, 108.5KB) *

clarkdave

#!/bin/bash -e

##
# Use this annotated script a base for launching an interactive console task on Amazon ECS
#
# more info: https://engineering.loyaltylion.com/running-an-interactive-console-on-amazon-ecs-c692f321b14d
#
# Requirements:
# - `jq` must be installed on both the client and server
##