Official release of code for the paper RL is a hammer and LLMs are nails A simple RL approach to stronger prompt injection attacks - View it on GitHub