Skill-Vetter-ZH 中文
安装任何 AI 智能体技能前的安全审查协议。检测凭据窃取、混淆代码、数据外泄等危险标志。风险分级:低危/中危/高危/极危。生成结构化审查报告。在未经审查的情况下,禁止安装任何来源不明的技能。 Skill: Skill-Vetter-ZH 中文 Owner: xocio Summary: 安装任何 AI 智能体技能前的安全审查协议。检测凭据窃取、混淆代码、数据外泄等危险标志。风险分级:低危/中危/高危/极危。生成结构化审查报告。在未经审查的情况下,禁止安装任何来源不明的技能。 Tags: latest:1.0.0 Version history: v1.0.0 | 2026-05-02T05:53:23.411Z | user Initial release with comprehensive security vetting protocol for AI skills: - Detects dangerous patterns like credential theft, code obfuscation, and data exfiltration. - Provides a systematic pre-in
Rank
62
Safety
84
Downloads
2.1k
Updated
Oct 9, 2026
Version
1.0.0
Source
CLAWHUB
About
What it does, and when to use it.
Capability contract not published. No trust telemetry is available yet. 2.1K downloads reported by the source. Last updated 10/9/2026.
Avoid when
- Contract metadata is missing or unavailable for deterministic execution.
Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing
Public facts
Every fact links back to the source it came from.
- Vendor
- Clawhubvendor · observed Oct 9, 2026
- Protocol compatibility
- OpenClawcompatibility · observed Oct 9, 2026
- Adoption signal
- 2.1K downloadsadoption · observed Oct 9, 2026
- Latest release
- 1.0.0release · observed May 2, 2026
- Handshake status
- UNKNOWNsecurity
Install and run
Setup complexity: low.
clawhub skill install s172j86vyw5s0s30q1k781zkd584w8w9:skill-vetter-zh- Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.
- Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.
Contract: missing
curl -s "https://www.xpersona.co/api/v1/agents/clawhub-xocio-skill-vetter-zh/snapshot"
Documentation
CLAWHUB
9,604 characters of source documentation, loaded on request.
Extracted files
4 files captured from the source.
SKILL.md
---
name: skill-vetter-zh
version: 1.0.0
description: 安装任何 AI 智能体技能前的安全审查协议。检测凭据窃取、混淆代码、数据外泄等危险标志。风险分级:低危/中危/高危/极危。生成结构化审查报告。在未经审查的情况下,禁止安装任何来源不明的技能。
homepage: https://clawhub.com
changelog: 初始版本 - 来源检查、代码审查清单、危险标志检测、权限分析、风险分级、审查报告模板
metadata:
openclaw:
emoji: "🔒"
requires:
bins: ["curl", "jq"]
os:
- linux
- darwin
- win32
---
# 技能审查器(Skill Vetter)🔒
为 AI 智能体技能提供以安全为核心的审查协议。**在安装任何技能之前,必须先完成审查。**
## 解决的问题
安装不可信的技能存在严重安全风险:
- 恶意代码可能窃取凭据
- 技能可能将数据外泄至外部服务器
- 混淆脚本可能执行任意命令
- 仿冒名称可能诱骗你安装假冒技能
本技能提供系统化的安装前审查流程。
## 适用场景
- **从 ClawHub 安装任何技能之前**
- **从 GitHub 仓库运行技能之前**
- **评估其他智能体分享的技能时**
- **收到安装未知代码请求时**
## 审查协议
### 第一步:来源检查
回答以下问题:
- [ ] 该技能来自哪里?
- [ ] 作者是否已知且可信?
- [ ] 下载量或星级如何?
- [ ] 最近一次更新是什么时候?
- [ ] 是否有其他智能体的使用评价?
### 第二步:代码审查(必做)
**阅读技能中的所有文件**,检查以下**危险标志**:
```
🚨 发现以下任一情况,立即拒绝:
─────────────────────────────────────────
• 向未知 URL 发起 curl/wget 请求
• 向外部服务器发送数据
• 请求凭据、令牌或 API 密钥
• 在没有明确原因的情况下访问 ~/.ssh、~/.aws、~/.config
• 访问 MEMORY.md、USER.md、SOUL.md、IDENTITY.md
• 对任何内容使用 base64 解码
• 对外部输入使用 eval() 或 exec()
• 修改工作区以外的系统文件
• 在未列出依赖的情况下安装软件包
• 向 IP 地址(而非域名)发起网络调用
• 代码混淆(压缩、编码、最小化)
• 请求 sudo 或 root 提升权限
• 访问浏览器 Cookie 或会话信息
• 接触凭据文件
─────────────────────────────────────────
```
### 第三步:权限范围评估
评估以下内容:
- [ ] 需要读取哪些文件?
- [ ] 需要写入哪些文件?
- [ ] 会执行哪些命令?
- [ ] 是否需要网络访问?访问哪里?
- [ ] 权限范围是否与其声明的用途相符且最小化?
**最小权限原则:** 技能应仅访问其绝对必要的资源。
### 第四步:风险分级
| 风险等级 | 典型示例 | 处理方式 |
|----------|----------|----------|
| 🟢 **低危** | 笔记、天气、文本格式化 | 基本审查后可安装 |
| 🟡 **中危** | 文件操作、浏览器、API 调用 | 需完整代码审查 |
| 🔴 **高危** | 凭据操作、交易、系统修改 | 需用户明确批准 |
| ⛔ **极危** | 安全配置、root 访问 | 禁止安装 |
## 审查清单(可直接复制使用)
```markdown
## 技能审查报告 — [技能名称] v[版本号]
**日期:** [日期]
**来源:** [URL]
**审查方:** [智能体名称]
### 自动化检查
- [ ] 无对用户可控输入的 exec 调用
- [ ] 无向未知域名发起的出站网络请求
- [ ] 无凭据收集模式
- [ ] 无工作区以外的文件系统访问
- [ ] 依赖项锁定至具体版本
- [ ] 无混淆或最小化代码
### 人工检查
- [ ] 作者有发布历史(非全新账号)
- [ ] 下载量与发布时间相符
- [ ] README 清楚说明了技能的实际功能
- [ ] 无"相信我"或制造紧迫感的措辞
- [ ] 更新日志存在且内容合理
### 结论
**风险等级:** 低危 / 中危 / 高危
**建议:** 可安装 / 谨慎安装 / 禁止安装
**备注:** [具体关注点]
```
## 审查报告模板
完成审查后,请生成以下格式的报告:
```
技能审查报告
═══════════════════════════════════════
技能名称:[name]
来源:[ClawHub / GitHub / 其他]
作者:[用户名]
版本:[version]
───────────────────────────────────────
统计数据:
• 下载量/星级:[数量]
• 最近更新:[日期]
• 已审查文件数:[数量]
───────────────────────────────────────
危险标志:[无 / 列出详情]
所需权限:
• 文件:[列表或"无"]
• 网络:[列表或"无"]
• 命令:[列表或"无"]
───────────────────────────────────────
风险等级:[🟢 低危 / 🟡 中危 / 🔴 高危 / ⛔ 极危]
审查结论:[✅ 可安全安装 / ⚠️ 谨慎安装 / ❌ 禁止安装]
备注:[其他观察说明]
═══════════════════════════════════════
```
## 快速审查命令
针对托管在 GitHub 上的技能:
```bash
# 查看仓库统计信息
curl -s "https://api.github.com/repos/OWNER/REPO" | \
jq '{stars: .stargazers_count, forks: .forks_count, updated: .updated_at}'
# 列出技能文件
curl -s "https://api.github.com/repos/OWNER/REPO/contents/skills/SKILL_NAME" | \
jq '.[].name'
# 获取并审查 SKILL.md
curl -s "https://raw.githubusercontent.com/OWNER/REPO/main/skills/SKILL_NAME/SKILL.md"
```
针对 ClawHub 上的技能:
```bash
# 搜索并查看热度信息
clawhub search "skill-name"
README.md
# 源项目 https://clawhub.ai/spclaudehome/skill-vetter # Skill Vetter(技能审查器)- 安装说明 为 AI 智能体技能提供以安全为核心的审查协议。 ## 快速安装 ```bash # 通过 ClawHub 安装(发布后可用) clawhub install skill-vetter # 或手动安装 cd ~/.openclaw/workspace/skills # 从 ClawHub 下载或解压安装包 ``` ## 使用方法 在安装**任何**技能之前: ``` 你:"请审查 ClawHub 上的 deep-research-pro 技能" 助手:[下载至临时目录,审查代码,检查危险标志] 助手:[生成包含风险等级和审查结论的报告] ``` ## 检查内容 - ✅ 来源信誉(下载量、星级、作者背景) - ✅ 代码审查(危险标志检测) - ✅ 权限范围(文件、网络、命令访问) - ✅ 风险分级(低危 / 中危 / 高危 / 极危) ## 危险标志(自动拒绝) - 向未知 URL 发起 curl/wget 请求 - 尝试窃取凭据或 API 密钥 - 混淆代码(base64 编码、压缩混淆、代码最小化) - 请求 sudo/root 权限 - 向 IP 地址发起网络调用 - 访问 ~/.ssh、~/.aws 等敏感目录 ## 审查报告示例 ``` 技能审查报告 ═══════════════════════════════════════ 技能名称:example-skill 来源:ClawHub 危险标志:无 所需权限:仅读写工作区 风险等级:🟢 低危 审查结论:✅ 可安全安装 ═══════════════════════════════════════ ``` ## 集成说明 **可与以下技能配合使用:** - **zero-trust-protocol(零信任协议):** 在审查流程中强制执行验证 - **drift-guard(偏移守卫):** 记录审查决策日志 ## 系统要求 - `curl`(用于 GitHub API 查询) - `jq`(用于 JSON 解析) ## 许可证 MIT - 可自由使用、修改、分发。 --- **永远不要安装未经审查的代码。先审查,后安装。**
_meta.json
{
"ownerId": "kn77ww7gabc6c6tqz2mqs0m8y984x0j6",
"slug": "skill-vetter-zh",
"version": "1.0.0",
"publishedAt": 1777701203411
}skill-card.md
## Description: A Chinese-language pre-installation review protocol for evaluating AI agent skills for credential theft, obfuscation, data exfiltration, permission scope, and installation risk. This skill is ready for commercial/non-commercial use. ## Publisher: [xocio](https://clawhub.ai/user/xocio) ### License/Terms of Use: MIT-0 ## Use Case: Developers, security reviewers, and agent operators use this skill before installing or running third-party skills to inspect source, permissions, network behavior, and risk indicators, then produce a structured review report. ### Deployment Geography for Use: Global ## Known Risks and Mitigations: Risk: Manual checklist results can miss issues if the reviewer skips files or relies only on source reputation. Mitigation: Read all files in the candidate skill, compare requested permissions to the stated purpose, and pair the manual review with automated scans where available. Risk: Example shell commands may fetch remote content or print local files when copied without adapting the target path or repository. Mitigation: Run review commands only against a temporary copy of the candidate skill or its public repository, and inspect domains and paths before execution. ## Reference(s): - [ClawHub skill page](https://clawhub.ai/xocio/skills/skill-vetter-zh) - [README project link](https://clawhub.ai/spclaudehome/skill-vetter) ## Skill Output: **Output Type(s):** [guidance, markdown, shell commands] **Output Format:** [Markdown review checklist and structured security report with optional shell command examples] **Output Parameters:** [1D] **Other Properties Related to Output:** [Chinese-language review workflow with risk grading and install recommendations] ## Skill Version(s): 1.0.0 (source: frontmatter and server release metadata) ## Ethical Considerations: Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.
AionUi
Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!
activepieces
AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents
cherry-studio
AI productivity studio with smart chat, autonomous agents, and 300+ assistants.
CopilotKit
The Frontend for Agents & Generative UI. React + Angular
Machine-readable data
The same record, as JSON, for agents and crawlers.
{
"facts": [
{
"factKey": "vendor",
"category": "vendor",
"label": "Vendor",
"value": "Clawhub",
"href": "https://clawhub.ai/xocio/skills/skill-vetter-zh",
"sourceUrl": "https://clawhub.ai/xocio/skills/skill-vetter-zh",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-09T18:34:50.403Z",
"isPublic": true
},
{
"factKey": "protocols",
"category": "compatibility",
"label": "Protocol compatibility",
"value": "OpenClaw",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-xocio-skill-vetter-zh/contract",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-xocio-skill-vetter-zh/contract",
"sourceType": "contract",
"confidence": "medium",
"observedAt": "2026-10-09T18:34:50.403Z",
"isPublic": true
},
{
"factKey": "traction",
"category": "adoption",
"label": "Adoption signal",
"value": "2.1K downloads",
"href": "https://clawhub.ai/xocio/skill-vetter-zh",
"sourceUrl": "https://clawhub.ai/xocio/skill-vetter-zh",
"sourceType": "profile",
"confidence": "medium",
"observedAt": "2026-10-09T18:34:50.403Z",
"isPublic": true
},
{
"factKey": "latest_release",
"category": "release",
"label": "Latest release",
"value": "1.0.0",
"href": "https://clawhub.ai/xocio/skill-vetter-zh",
"sourceUrl": "https://clawhub.ai/xocio/skill-vetter-zh",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-05-02T05:53:23.411Z",
"isPublic": true
},
{
"factKey": "handshake_status",
"category": "security",
"label": "Handshake status",
"value": "UNKNOWN",
"href": "https://www.xpersona.co/api/v1/agents/clawhub-xocio-skill-vetter-zh/trust",
"sourceUrl": "https://www.xpersona.co/api/v1/agents/clawhub-xocio-skill-vetter-zh/trust",
"sourceType": "trust",
"confidence": "medium",
"observedAt": null,
"isPublic": true
}
],
"events": [
{
"eventType": "release",
"title": "Release 1.0.0",
"description": "Initial release with comprehensive security vetting protocol for AI skills: - Detects dangerous patterns like credential theft, code obfuscation, and data exfiltration. - Provides a systematic pre-installation review checklist and risk grading (low/medium/high/critical). - Includes structured audit report templates for consistent decision-making. - Enforces a policy: Skills from unknown sources cannot be installed without prior review. - Supports command-line review procedures for both ClawHub and GitHub sources.",
"href": "https://clawhub.ai/xocio/skill-vetter-zh",
"sourceUrl": "https://clawhub.ai/xocio/skill-vetter-zh",
"sourceType": "release",
"confidence": "medium",
"observedAt": "2026-05-02T05:53:23.411Z",
"isPublic": true
}
]
}Record generated Oct 9, 2026.
