Skill Scanner
Use snyk/agent-scan to detect security risks in agent components.
Quick Scan
CODEBLOCK0
What It Detects
For Skills
- - Prompt Injection (E004) - Malicious instructions hidden in prompts
- Malware Payloads (E006) - Harmful code disguised as content
- Untrusted Content (W011) - Potentially unsafe external data
- Credential Handling (W007) - Improper secrets management
- Hardcoded Secrets (W008) - API keys or passwords in code
For MCP Servers
- - Prompt Injection (E001)
- Tool Poisoning (E003)
- Tool Shadowing (E002)
- Toxic Flows (TF001)
- Rug Pull (W005) - Malicious skill replacement
Workflow
- 1. Before installing a new skill → Run a scan first
- After scanning → Review any E001/E003/E004/E006 issues (high severity)
- Low severity warnings (W005-W008) → Decide based on your risk tolerance
Interpreting Results
| Prefix | Severity | Action |
|---|
| E | High | Fix or avoid the skill |
| W |
Medium/Low | Review and decide |
|
TF | High | Toxic flow detected |
Common Issues
If uvx is not found, install uv first:
CODEBLOCK1
OpenClaw Skills Location
OpenClaw skills are typically stored at:
- - Global: INLINECODE1
- Workspace: INLINECODE2
To scan a custom path, pass it directly:
CODEBLOCK2
Output Example
The scan will show:
- - File path of the issue
- Risk type and description
- Severity level (E/W/TF)
- Recommended fix
Review the full report at: https://github.com/snyk/agent-scan/blob/main/docs/issue-codes.md
技能扫描器
使用 snyk/agent-scan 检测代理组件中的安全风险。
快速扫描
bash
扫描机器上的所有技能
uvx snyk-agent-scan@latest --skills
扫描 MCP 服务器（默认行为）
uvx snyk-agent-scan@latest
带详细输出的扫描
uvx snyk-agent-scan@latest --skills --verbose
输出 JSON 格式用于自动化
uvx snyk-agent-scan@latest --skills --json
检测内容
针对技能
- - 提示注入 (E004) - 隐藏在提示中的恶意指令
- 恶意软件载荷 (E006) - 伪装成内容的有害代码
- 不可信内容 (W011) - 潜在不安全的的外部数据
- 凭证处理 (W007) - 不当的机密管理
- 硬编码机密 (W008) - 代码中的 API 密钥或密码
针对 MCP 服务器
- - 提示注入 (E001)
- 工具投毒 (E003)
- 工具影子攻击 (E002)
- 毒性流程 (TF001)
- 拉地毯骗局 (W005) - 恶意技能替换
工作流程
- 1. 安装新技能前 → 先运行扫描
- 扫描后 → 检查所有 E001/E003/E004/E006 问题（高严重性）
- 低严重性警告 (W005-W008) → 根据您的风险承受能力决定
结果解读
中/低 | 审查并决定 |
|
TF | 高 | 检测到毒性流程 |
常见问题
如果找不到 uvx，请先安装 uv：
bash
macOS
brew install uv
Linux
curl -LsSf https://astral.sh/uv/install.sh | sh
OpenClaw 技能存储位置
OpenClaw 技能通常存储在：
- - 全局: ~/.openclaw/skills/
- 工作区: /skills/
要扫描自定义路径，直接传入即可：
bash
uvx snyk-agent-scan@latest ~/.openclaw/skills/
输出示例
扫描将显示：
- - 问题的文件路径
- 风险类型和描述
- 严重性级别 (E/W/TF)
- 推荐的修复方案
查看完整报告：https://github.com/snyk/agent-scan/blob/main/docs/issue-codes.md
