Skills for building and controlling physics simulations with MuJoCo. These skills provide comprehensive reference documentation and agent tooling for programmatic control of MuJoCo simulations and humanoid robots.
- Docs: https://mujoco.readthedocs.io/en/stable/python.html (Python bindings API reference)
- Models: https://github.com/google-deepmind/mujoco_menagerie (ready-made robot models)
| Skill | Description |
|---|---|
| mujoco-python | Python bindings reference — loading models, stepping physics, joint control, sensor reading, camera control, offscreen rendering, visualization flags/groups, and solver configuration |
npx skills add prathje/agentic_mujoco_skillsContributions welcome! If you find incorrect documentation, missing API coverage, or want to add new skills (e.g., MJX, dm_control, humanoid locomotion), please open a PR.