Can Large Language Models Help Students Prove Software Correctness?