Robotics Conventions
Use for robotics / robot-learning code when handling frames, units, axes, signs, quaternions, URDFs, sim-to-real, RL/IL policies, kinematics, hardware, visualization, or data pipelines.
Install
$ agentstack add skill-berkeley-humanoids-skills-robotics-conventions ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Links to your public security report.[](https://agentstack.voostack.com/security/report/skill-berkeley-humanoids-skills-robotics-conventions) Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Robotics conventions
- Prefer explicit names in variables, methods, classes, and configs.
- Use
_left/_right, not_l/_r. - Avoid ambiguous abbreviations unless they are standard in robotics.
- Use SI units everywhere: meters, seconds, radians, kilograms, Newtons, and Newton-meters.
- Non-SI values are only allowed at third-party, UI, logging, or hardware API boundaries.
- Mark non-SI variables with suffixes such as
_deg,_ms,_mm.
- Use scalar-first quaternion order:
(qw, qx, qy, qz). - Normalize quaternions before storing or applying them.
- Convert explicitly at library/message boundaries when another order is required.
- Use REP-103 robot frames:
xforward,yleft,zup, right-handed. - Blender and Onshape use
yforward,zup. - Prefer realigning CAD assets or motion data during export, not patching frame conversions in runtime code.
- Training environment setup should follow this order:
CommandObservationActionMetricsRewardsTerminationsEventsCurriculum
- Robot joint ordering should follow a depth-first body traversal.
- Example: left arm shoulder-to-hand, right arm shoulder-to-hand, left leg hip-to-foot, right leg hip-to-foot.
- Keep the ordering defined in one source of truth and reference it everywhere.
- Use and reference ROS 2 documentation, not ROS 1 documentation.
- Prefer environment managers:
uvfor Python environments.pixifor C/C++ and ROS 2 environments.
- Preferred data formats:
.mcapfor raw data recording, especially when interfacing with C/C++ code.LeRobotDataset/.parquetfor training datasets stored as Hugging Face dataset repositories..rblfor saved Rerun visualizations that combine.mcapdata with a Rerun layout.
- Preferred visualization tools:
- Rerun for multimodal data visualization, especially require custom Blueprint panels.
- Viser for only motion and pose visualization.
- MuJoCo for physics-based interactive visualization.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: Berkeley-Humanoids
- Source: Berkeley-Humanoids/Skills
- License: CC0-1.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.