Minimax-Optimal Multi-Agent Robust Reinforcement Learning