XPolicyLab

XPolicyLab is a unified standard and open ecosystem for robot policy evaluation and deployment, released in August 2026 by the XPolicyLab Community. It specifies common observation, action, and trajectory schemas together with a minimal adapter interface for observation updates, action prediction, batched execution, and episode reset. A dependency-isolated client/server architecture separates policy inference from environment execution, allowing each side to retain its native software stack and run locally or remotely. This reduces the cost of connecting N policies to M evaluation environments from O(NM) to O(N+M). As of August 2026, the ecosystem integrates 42 robot policies spanning VLA, world-action, imitation-learning, and memory-augmented families. The same adapters serve RoboTwin and RoboDojo simulation as well as standardized real-robot evaluation through one interface; the official RoboDojo and RoboTwin leaderboards are powered by XPolicyLab submissions. In a controlled study, conforming to the standard reduced the integration effort of a representative policy from over five hours to two hours, and to thirty minutes with packaged agent skills. Code is released on GitHub at github.com/XPolicyLab/XPolicyLab, with the project website at xpolicylab.github.io.