← 返回 Skills 市场
1888
总下载
2
收藏
3
当前安装
2
版本数
在 OpenClaw 中安装
/install vector-control
功能描述
Control a Vector robot via Wirepod’s local HTTP API on the same network. Use when you need to move Vector, tilt head/lift, speak text, capture camera frames, or run patrol/explore routines from the Pi/Wirepod host. Includes a CLI helper script and endpoint reference.
安全使用建议
This skill appears to do what it says: control a Vector robot over the local Wirepod HTTP API. Before installing/using it: (1) Ensure ffmpeg is installed on the host (the script calls ffmpeg but the skill metadata doesn't declare it). (2) Run the script on a trusted local network — the --base argument can point the script to any URL, so avoid directing it to untrusted external hosts (this would let audio or control commands go off-network). (3) When using the play command, only provide audio files you trust; the script reads them and writes a temporary WAV in /tmp which it removes afterwards. (4) Optionally inspect the script yourself (it's included) to confirm behavior before use. If you need the skill to be restricted to localhost only, consider modifying the script to enforce base='http://localhost:8080' or validating --base at runtime.
功能分析
Type: OpenClaw Skill
Name: vector-control
Version: 1.0.1
The skill is designed to control a local robot via a local API. While the `SKILL.md` instructions are benign and do not contain prompt injection attempts, the `scripts/vector_control.py` script includes risky capabilities. Specifically, it uses `subprocess.run` to execute `ffmpeg` with user-controlled input file paths (`args.file`), and allows writing camera snapshots to arbitrary local paths (`args.out`). Although these capabilities are plausibly needed for the stated purpose (audio conversion, saving snapshots), they represent broad file system access and command execution that could be leveraged by a malicious prompt to the AI agent, even if the skill itself does not explicitly instruct such harmful actions. All network communication is confined to `http://localhost:8080`, indicating no external data exfiltration.
能力评估
Purpose & Capability
The skill's name/description align with the included script and API reference: it implements movement, speech, camera snapshot, patrol/explore, and audio playback via Wirepod HTTP endpoints. One minor mismatch: the script invokes ffmpeg to convert audio but the SKILL.md / registry metadata do not declare ffmpeg or any required binary dependency.
Instruction Scope
SKILL.md instructs the agent/user to run the bundled CLI script and to use local Wirepod endpoints (default base http://localhost:8080). The script itself reads only user-supplied audio files and writes temporary WAVs to /tmp, and uses the Wirepod API. It does allow overriding --base to any URL, which could direct requests (including audio uploads) to a remote host if misused — the documentation expects same-network/local usage but does not enforce it.
Install Mechanism
There is no install spec (instruction-only plus a script), so nothing will be automatically downloaded or written by an installer. This is low-risk. Note: runtime dependency on ffmpeg is present but not declared; ffmpeg must be available on PATH for audio playback to work.
Credentials
The skill does not request any environment variables, credentials, or configuration paths. The SKILL.md references a local file path where an ESN may be found (/etc/wire-pod/.../botSdkInfo.json), but the script does not automatically read arbitrary system credentials. File I/O is limited to user-specified audio files and temporary /tmp files.
Persistence & Privilege
The skill is not always-enabled and does not request persistent system privileges or attempt to modify other skills or system-wide agent settings. It runs as a user-invoked CLI and performs network calls only to the configured base URL.
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 vector-control - 安装完成后,直接呼叫该 Skill 的名称或使用
/vector-control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
v1.0.1
Add patrol/explore, audio streaming, camera snapshots
v1.0.0
Initial public release of vector-control.
- Provides local control of a Vector robot via the Wirepod HTTP API.
- Includes CLI script for movement, head/lift control, speech, camera snapshots, and audio playback.
- Supports patrol and exploration routines directly from the Pi/Wirepod host.
- Documents all key endpoints and usage instructions.
- Offers full reference for Wirepod’s API and control tasks.
元数据
常见问题
Vector Control 是什么?
Control a Vector robot via Wirepod’s local HTTP API on the same network. Use when you need to move Vector, tilt head/lift, speak text, capture camera frames, or run patrol/explore routines from the Pi/Wirepod host. Includes a CLI helper script and endpoint reference. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1888 次。
如何安装 Vector Control?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install vector-control」即可一键安装,无需额外配置。
Vector Control 是免费的吗?
是的,Vector Control 完全免费(开源免费),可自由下载、安装和使用。
Vector Control 支持哪些平台?
Vector Control 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 Vector Control?
由 dbeadle1(@dbeadle1)开发并维护,当前版本 v1.0.1。
推荐 Skills