game-environments is an open-source package offering standalone game theory environments for evaluating and benchmarking AI agents. It supports scenarios like the prisoners' dilemma and nash equilibrium, providing a CLI for easy integration into research workflows. Ideal for AI researchers and developers focused on agent evaluation.
game-environments sits in PulseGate's Agent evaluation & testing category. It enables researchers and developers to evaluate AI agents in standardized game theory environments. It is built as an open-source project for AI researchers and developers. game-environments is open source under the MIT License. It ships for the command line, and it can be self-hosted.
It is developed by atp-platform, and it first shipped in 2026. Among its 5 catalogued features are game theory environments, agent evaluation, and nash equilibrium support.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do