← 返回 Skills 市场
214
总下载
1
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install robotics-vla
功能描述
Expert guidance for Vision-Language-Action (VLA) robot foundation models — covering architecture design, training pipelines, data strategy, deployment, and e...
安全使用建议
This skill is a documentation bundle offering technical guidance for building and training VLA robot models; it does not request credentials or install code and appears internally consistent. However, the source is unknown and claims (model sizes, datasets, performance) are not independently verifiable from the package — treat the content as expert guidance, not a tested implementation. Before using any recommendations in production or on hardware: (1) verify provenance of referenced models/data, (2) cross-check the training/operation details against primary literature or open-source implementations, and (3) perform safety testing in simulation to avoid risking real robots. If you need higher assurance, ask the publisher for provenance, training logs, or reproducible scripts.
功能分析
Type: OpenClaw Skill
Name: robotics-vla
Version: 1.1.0
The skill bundle contains purely informational and technical documentation regarding Vision-Language-Action (VLA) robotics models, specifically the π0 architecture. There is no executable code, scripts, or suspicious instructions; the content is limited to architectural details, training strategies, and references to robotics research (SKILL.md, architecture.md, training.md).
能力评估
Purpose & Capability
The skill's name and description match the included SKILL.md and reference documents: architecture, training, embodiments, and related work for a VLA (π0) style robot model. No unrelated binaries, credentials, or config paths are requested.
Instruction Scope
Runtime instructions are purely advisory (design/training/evaluation guidance) and only reference bundled docs. There are no instructions to read system files, environment variables, network endpoints, or to perform actions outside the documentation scope.
Install Mechanism
No install specification or code files are present—this is instruction-only material, so nothing is downloaded or written to disk by an installer.
Credentials
The skill declares no required environment variables, credentials, or config paths. There are no extra secret-like variables requested that would be disproportionate to its stated purpose.
Persistence & Privilege
The skill does not request always:true and is user-invocable with normal autonomy settings. It does not attempt to modify other skills or system settings; no persistence or elevated privileges are requested.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install robotics-vla - 安装完成后，直接呼叫该 Skill 的名称或使用
/robotics-vla触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.1.0
Add 2025 research landscape: pi0-FAST/0.5/0.6 successors, RTC async chunking, UniVLA unsupervised actions, ManiFlow, GR00T N1, Helix, OpenVLA-OFT
v1.0.0
Initial release: pi0 VLA architecture, flow matching, training pipeline, multi-embodiment support
元数据
常见问题
Robotics VLA 是什么？
Expert guidance for Vision-Language-Action (VLA) robot foundation models — covering architecture design, training pipelines, data strategy, deployment, and e... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 214 次。
如何安装 Robotics VLA？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install robotics-vla」即可一键安装，无需额外配置。
Robotics VLA 是免费的吗？
是的，Robotics VLA 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Robotics VLA 支持哪些平台？
Robotics VLA 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 Robotics VLA？
由 arden2010（@arden2010）开发并维护，当前版本 v1.1.0。
推荐 Skills