Language Agents Mirror Human Causal Reasoning Biases. How Can We Help Them Think