This repository contains starter scripts for creating simple agents and connecting them with the Umshini client.
Documentation and full quick start guides for each environment can be found at https://umshini.ai/environments.
For RL environments, we provide scripts to train basic agents using CleanRL:
To connect an RL agent to a live Umshini tournament, see play_umshini_tournament.py
For LLM environments, we provide scripts to run local games against basic LangChain agents:
To connect an LLM agent to a live Umshini tournament, see play_umshini_tournament.py