Safe Multi-Agent Robosuite benchmark for safe multi-agent reinforcement learning research. - View it on GitHub