Check out my agent implementation here which is just for fun and to get a better understanding. Maybe you want to contribute?

https://github.com/mpaepper/llm_agents