Description
This project focuses on designing and implementing evaluations for AI coding agents to ensure they adhere to ethical guidelines and maintain high standards of conduct. The primary objective is to assess whether AI agents perform tasks safely, accurately, and responsibly rather than merely completing them efficiently. The role involves creating challenging scenarios where the optimal solution might be counterintuitive, followed by rigorous testing to identify deviations from expected behavior.
