We present RLxUSD v0.1, a minimal and extensible convention to encode complete reinforce-ment learning (RL) episodes as OpenUSD scenes. The convention prescribes a small set of sceneprimitives, a metrics: namespace for time-aligned time series, and an episode summary storedin stage metadata with both required and optional fields. By leveraging USD timeSamples andnative extensibility, RLxUSD unifies geometry, metrics, and metadata in a single, inspectableartifact. We provide a compact reference implementation, integrations with Gymnasium andStable-Baselines3 (SB3), a usdview heads-up display (HUD) for synchronized visualization, anda small reproducible dataset across Random, Greedy, and PPO agents in 16×16 and 32×32gridworlds. The result improves transparency, portability, and reproducibility for episode-levelRL research and engineering.