Environment Hub
A shared catalog of RL training environments — tasks bundled with their verifiers — that agents train against.
An environment hub is a repository of reinforcement-learning environments: each entry packages a task, the tools the agent may use, and a verifier that scores success. Instead of every lab rebuilding tasks, a hub lets environments be shared and composed, making the environment — not just the algorithm — the thing that teaches an agent.