← 返回 Skills 市场
17863
总下载
9
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install skill-vetting
功能描述
Vet ClawHub skills for security and utility before installation. Use when considering installing a ClawHub skill, evaluating third-party code, or assessing w...
安全使用建议
Install only if you want a lightweight local checklist and regex scanner, and treat its output as advisory. Use a private temporary directory, validate ZIP contents before extraction, and do not let the skill's automatic-reject rules replace contextual review.
功能分析
Type: OpenClaw Skill
Name: skill-vetting
Version: 1.1.0
This skill bundle is a security scanner designed to vet other OpenClaw skills for malicious patterns and prompt injection. The `SKILL.md` and `ARCHITECTURE.md` files explicitly warn against prompt injection and provide instructions for the AI agent on how to identify and reject such attempts. The `scripts/scan.py` file implements the detection logic for various security risks, including prompt injection, but does not exhibit any malicious behavior itself. All files are aligned with the stated purpose of security vetting and show no evidence of intentional harmful behavior, data exfiltration, unauthorized execution, or persistence mechanisms.
能力评估
Credentials
Network, shell, and file-read capabilities are expected for reviewing skill packages, but they are documented as raw commands without narrow permission declarations or clear containment controls.
Install Mechanism
No install-time payload or dependency install was found, but the documented workflow downloads a remote ZIP and extracts it with unzip in /tmp without archive path, size, signature, or unique-directory validation.
Instruction Scope
SKILL.md tells the agent that scanner findings are ground truth, never to downgrade them, and to automatically reject some findings; those review-control instructions are overbroad even though the defensive goal is coherent.
Persistence & Privilege
I found no persistent service, startup hook, credential harvesting, privilege escalation, or background worker in the shipped artifacts; flagged credential and destructive strings are inert examples.
Purpose & Capability
The stated purpose matches the included local regex scanner and review checklist, but the scanner is context-insensitive and the skill treats its findings too authoritatively for a final security decision.
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 skill-vetting - 安装完成后，直接呼叫该 Skill 的名称或使用
/skill-vetting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
Security: Expanded file scanning (html/env/ts/etc) + 5 new prompt injection patterns. Features: JSON output, severity levels, color-coded findings. Docs: AI reviewer hardening + defense architecture.
v1.0.1
- Added SKILL.md.backup file for backup or versioning purposes.
- No changes to core functionality or documentation content.
- The skill continues to provide security and utility vetting guidance for ClawHub skills.
v1.0.0
Initial release - Automated security scanner + malicious pattern database + complete vetting workflow for ClawHub skills
元数据
常见问题
Skill Vetting 是什么？
Vet ClawHub skills for security and utility before installation. Use when considering installing a ClawHub skill, evaluating third-party code, or assessing w... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 17863 次。
如何安装 Skill Vetting？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install skill-vetting」即可一键安装，无需额外配置。
Skill Vetting 是免费的吗？
是的，Skill Vetting 完全免费（开源免费），可自由下载、安装和使用。
Skill Vetting 支持哪些平台？
Skill Vetting 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 Skill Vetting？
由 Eddy（@eddygk）开发并维护，当前版本 v1.1.0。
推荐 Skills