Hypothetical Minds: Scaffolding Theory of Mind for Multi-Agent Tasks with Large Language Models