I thought it would be interesting to explore collaboration with LLMs, and love Codenames, so built out a fun way to play (without friends). It's running on gpt-3.5-turbo and sometimes has issues with "reasoning", so it might be worth trying this with gpt-4
