Arena is a new game playing framework, that I am developing in collaboration with Rem Collier (of the ASTRA language fame). We’re in fairly initial stages, but already it’s shaping up to be an exciting project. Think evolutionary computing meets agents meets game theory. We’ll put up more details as soon as we get some time to create some nice diagrams. In the meanwhile, take a look at our paper (preprint copy), that’s going to be presented at the Adaptive Learning Agents Workshop, co-located with AAMAS 2018.
Check out our git repo for the source code: https://gitlab.com/aop-arena
If you want to contribute, or just chat with us about its features, get in touch.