概要
This skill provides a systematic framework for optimizing MuJoCo MJCF models, guiding developers through the complex trade-offs between simulation speed and physical fidelity. It offers structured methodologies for distinguishing between integration discretization and constraint solver errors, enabling more effective parameter tuning across timesteps, solvers, and integrators. Whether you are balancing computation time for reinforcement learning or troubleshooting unstable trajectories, this skill ensures a methodical approach to simulation optimization through established baselines, strategic parameter sweeps, and rigorous verification protocols.
