MudGym

A Gymnasium reinforcement learning environment for MUD2. One of the first online multiplayer text adventure games.

Status

MudGym is a work in progress and very much under active development.

I've been using most of what's in here myself for a while, but the process of putting it out for wider use, and seeing how it gets used for real, means I'm still refining and settling some of the interfaces and assumptions.

So for now, be prepared for interfaces, APIs, specs and defaults to change, sometimes in breaking ways. If you're using MudGym, feedback and real-world use cases are very welcome.

Setup

Install MudGym from PyPI:

uv add mudgym

By default, MudGym now runs the MUD2 game engine via WASM, which is installed as a dependency.

Quickstart

from mudgym import make_env

env = make_env(observation="parsed")
observation, info = env.reset()
observation, reward, terminated, truncated, info = env.step("howl")
print(observation["room_name"], reward)
env.close()

Docs

See the docs for more examples.

Citing MudGym

If you use MudGym in your research, please cite it using the metadata in CITATION.cff.

Use Cite this repository in the GitHub sidebar to copy a citation in APA or BibTeX format. Please cite the version you used.

License

The Python code and tooling in this repository are made available under the MIT License.

Permission to use the MUD2 game for research purposes has been provided by Richard Bartle, with kind thanks. The MUD2 game, name, source code, and associated story remain the intellectual property of Richard Bartle and Roy Trubshaw and may not be used for commercial purposes.