This project is dedicated to analyzing security vulnerabilities in AI systems, focusing on agentic coding agents. Our goal is to raise awareness around critical risks like prompt injection and the dangers of over-reliance on LLM output.| monthofaibugs.com