ETHSecurity Community

For anyone interested. I did a deepdive on Open AI's codex security. TLDR: It's a small stack of skill files and some JS code that starts a large number of Codex sessions. They all get the prompt and the same scope with a limited set of tools. Not one prompt in the repo contains security advice, performance is obtained only through scaling number of agents looking at the code https://x.com/GustavHartz/status/2084926544800035266?s=20