扫描报告
0 /100
lobster-santa-method
Multi-agent adversarial verification with convergence loop. Two independent review agents must both pass before output ships.
A legitimate documentation-only skill describing a multi-agent adversarial verification framework. No executable code, network requests, credential access, or malicious patterns detected.
可以安装
This skill is safe to use. It provides a methodology for quality assurance of AI outputs using dual independent reviewers.
| 资源类型 | 声明权限 | 推断权限 | 状态 | 证据 |
|---|---|---|---|---|
| 文件系统 | NONE | NONE | — | No file operations referenced |
| 网络访问 | NONE | NONE | — | No network requests documented |
| 命令执行 | NONE | NONE | — | No shell execution in pseudocode patterns |
| 环境变量 | NONE | NONE | — | No env access documented |
| 技能调用 | READ | READ | ✓ 一致 | Uses Agent tool to spawn reviewers (documented as Pattern A) |
| 剪贴板 | NONE | NONE | — | No clipboard access |
| 浏览器 | NONE | NONE | — | No browser operations |
| 数据库 | NONE | NONE | — | No database access |
目录结构
1 文件 · 12.5 KB · 306 行 Markdown 1f · 306L
└─
SKILL.md
Markdown
安全亮点
✓ Documentation-only skill with no executable code
✓ No network requests or external communications
✓ No credential harvesting or sensitive data access
✓ No shell execution or command injection vectors
✓ No obfuscation techniques (base64, eval, etc.)
✓ Clear description of purpose and methodology
✓ No download or install components
✓ Pattern A references Claude Code Agent tool for subagent spawning - documented and intentional