LogoClawIndex
CasesSkillsAbout
LogoClawIndex

ClawIndex

OpenClaw Skills & Use Case Index

ClawIndex is an ecosystem-driven index of OpenClaw skills and real-world use cases.

Index

Skills·
Cases

Meta

About·
Disclaimer·
Email·
GitHub
© 2026 ClawIndex All Rights Reserved.

loco-manip-direct - Loco-Manip Direct Policy for G1 DEX3 Grasping

Executes G1 DEX3 hand policy to grasp, lift, and retain a tabletop cube before SIMPLE door pushing.

Tags

Updated: 2026-09-17

Capabilities

Typical Inputs

Typical Outputs

What this skill does

  • Grasp tabletop cube
  • Lift tabletop cube
  • Retain cube while standing
  • Relocalize object pose

Inputs

  • workspace.json
  • canonical SKILL.md
  • experience_read
  • Head and wrist RGB images
  • Robot measured and reference poses
  • Support plane calibration data

Outputs

  • scalebfm_act requests
  • scratch/NOTES.md
  • Rollout evidence descendants
  • Skill update submissions

Requirements

  • gpt-6-astra/xhigh model
  • Python environment at /home/galbot/lizekai/Loco_manip/ENV/locomanip_sim/bin/python
  • Registered physics services

Source

  • Spec: SKILL.md
loco-manipulation
g1-dex3
robotic-grasping
scalebfm
Grasp tabletop cube
Lift tabletop cube
Retain cube while standing
Relocalize object pose
workspace.json
canonical SKILL.md
experience_read
scalebfm_act requests
scratch/NOTES.md
Rollout evidence descendants