Mujoco

MuJoCo

This skill covers two MuJoCo task families. Do not load both detailed contexts at once.

First classify the user's task, then load only the relevant reference:

• MuJoCo Scene Builder: Create or modify MJCF / XML, build scenes, place robots/tables/objects, fix body / joint / geom / site / contact / inertial structure, compile/debug models, and open scenes for visual inspection.

Read [references/scene-builder.md](references/scene-builder.md).

• MuJoCo Robot Control: Inspect model structure, start a viewer, execute actuator control, run minimal control experiments, open/close grippers, and distinguish model problems from control-chain problems.

Read [references/robot-control.md](references/robot-control.md).

If a task includes both parts, use this fixed order:

1. Run Scene Builder first, and make the model or scene compile and open cleanly.
2. Run Robot Control second, start the viewer service, and send control commands.

Shared Rules

• Do not invent MuJoCo tags, attribute names, or default behaviors. If syntax is uncertain, use official docs and existing models in the current workspace as the source of truth.
• Do not guess user intent for scene creation or model edits. If the scene goal, robot choice, task semantics, object layout, physical constraints, output path, or validation expectation is unclear, ask a concise clarification question before creating or changing files.
• The default output directory is ~/Documents/mujoco. If the user does not provide an absolute path, prefer saving, searching, or resolving models there.
• When the user asks to "list all scenes", read only from ~/Documents/mujoco by default, and return only first-level directory names as scene names. Do not recursively expand files.
• When the user only says "open MuJoCo" or "open a scene.xml", infer intent first:
• If the goal is to check whether a scene opens normally, use Scene Builder.
• If the follow-up involves robot operation, also load Robot Control.
• When reporting results, always state the final model path, the command or script entry point you actually used, and any risk that remains unverified.

Official References

• MuJoCo modeling: https://mujoco.readthedocs.io/en/stable/modeling.html
• MuJoCo python: https://mujoco.readthedocs.io/en/stable/python.html
• MuJoCo api: https://mujoco.readthedocs.io/en/stable/APIreference/index.html

Scripts And Resources

• Viewer service entry point: scripts/mujoco_viewer.py
• Control and query entry point: scripts/mujoco_cli.py
• Scene physical sanity checker: scripts/mujoco_scene_check.py
• Grasp-chain verifier: scripts/mujoco_grasp_check.py
• Offscreen multi-view renderer: scripts/mujoco_render_views.py
• Environment and path helpers: scripts/env_bootstrap.py, scripts/path_utils.py
• Do not preload the entire scripts/ directory. Open a script only when you need to execute or modify it.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: coolbeevip
• Source: coolbeevip/mujoco-skills
• License: MIT

Install and usage instructions live in the source repository linked above.