Adds the capability to predict 6-DOF grasp poses for various robotic grippers. GraspGenX is a cross-embodiment model that can generate grasps for grippers unseen during training by conditioning on the gripper's swept volume.
Use this skill when you need to install GraspGenX, run inference on object point clouds or meshes, configure new grippers via the interactive wizard, or integrate grasping into an end-to-end pick-and-place pipeline using cuRobo and Newton/MuJoCo.
demo_object_pc.py), scene point clouds (demo_scene_pc.py), and object meshes (demo_object_mesh.py).Designed for agents with access to a GPU environment and terminal, compatible with MCP-enabled clients like Claude Desktop, Cursor, and VS Code.
This skill has not been reviewed by our automated audit pipeline yet.