As robots gain capabilities to enter our human-centric world, they require\nformalism and algorithms that enable smart and efficient interactions. This is\nchallenging, especially for robotic manipulators with complex tasks that may\nrequire collaboration with humans. Prior works approach this problem through\nreactive synthesis and generate strategies for the robot that guarantee task\ncompletion by assuming an adversarial human. While this assumption gives a\nsound solution, it leads to an "unfriendly" robot that is agnostic to the human\nintentions. We relax this assumption by formulating the problem using the\nnotion of regret. We identify an appropriate definition for regret and develop\nregret-minimizing synthesis framework that enables the robot to seek\ncooperation when possible while preserving task completion guarantees. We\nillustrate the efficacy of our framework via various case studies.\n