CJHwong

Claude Code Prompt PII Hook — moved

This project moved to github.com/CJHwong/agent-seatbelt under hooks-opf/.

The new version adds:

• PostToolUse blocking — catches PII in tool responses before the LLM sees them next turn, in addition to the prompt-side check.
• Codex support — auto-detects ~/.codex/ and wires hooks there too.
• Block-level tiers — PII_BLOCK_LEVEL=off|relaxed|standard|strict to choose what fires.
• Per-prompt bypass — prefix a prompt with pii:off to skip the check once.

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.

rob0tca

function extractTextOnOpen() {

    //ADD YOUR VALUES BELOW
    var folderName = "[YOUR PROJECT FOLDER]";
    var sheetId = "[YOUR SHEET ID]";

    //Define folder
    var folder = DriveApp.getFoldersByName(folderName).next();
    var folderId = folder.getId();