This is an early prototype of using prompting to create o1-like reasoning chains to improve output accuracy.