可信 — 风险评分 5/100
上次扫描:21 小时前 重新扫描
5 /100
genai-calling
Unified interface for all AI providers and modalities - text/image/audio/video/embedding workflows with MCP support
This is a documentation-only skill providing instructions for using the genai-calling CLI to interface with various AI model providers. No executable code, malicious patterns, or credential harvesting detected.
技能名称genai-calling
分析耗时28.9s
引擎pi
可以安装
No action required. This is a legitimate documentation skill. Consider pinning the genai-calling package version in production for reproducibility.

安全发现 1 项

严重性 安全发现 位置
低危
Package version not pinned 供应链
The skill uses 'uvx --from genai-calling' without specifying a version, which means it will always fetch the latest version from PyPI. This could introduce unexpected behavior if the package is updated.
uvx --from genai-calling genai
→ Pin the package version for reproducible behavior, e.g., 'uvx --from genai-calling==1.2.3 genai'
SKILL.md:31
资源类型声明权限推断权限状态证据
文件系统 NONE NONE No filesystem operations declared or inferred
网络访问 READ READ ✓ 一致 Skill documents connecting to AI provider APIs (expected behavior)
命令执行 WRITE WRITE ✓ 一致 Documents uvx CLI invocations, standard pattern
环境变量 READ READ ✓ 一致 Documents .env loading for API credentials (standard pattern)
技能调用 NONE NONE No skill chaining declared
剪贴板 NONE NONE No clipboard access documented
浏览器 NONE NONE No browser automation documented
数据库 NONE NONE No database access documented
4 项发现
🔗
中危 外部 URL 外部 URL
https://dashscope.aliyuncs.com/compatible-mode/v1
SKILL.md:122
🔗
中危 外部 URL 外部 URL
https://ark.cn-beijing.volces.com/api/v3
SKILL.md:125
🔗
中危 外部 URL 外部 URL
https://api.tu-zi.com
SKILL.md:127
🔗
中危 外部 URL 外部 URL
https://api.tu-zi.com/v1
SKILL.md:128

目录结构

1 文件 · 9.6 KB · 322 行
Markdown 1f · 322L
└─ 📝 SKILL.md Markdown 322L · 9.6 KB

依赖分析 1 项

包名版本来源已知漏洞备注
genai-calling latest (unpinned) PyPI (uvx) Version not pinned - will always fetch latest

安全亮点

✓ Documentation-only skill with no executable code present
✓ No credential harvesting - only documents standard env var loading
✓ No base64, obfuscation, or anti-analysis patterns detected
✓ No sensitive path access (SSH, AWS configs, .env files)
✓ No reverse shell, C2, or data exfiltration patterns
✓ External URLs are legitimate AI provider endpoints (Aliyun, Volcengine, Tu-zi)
✓ Standard package runner (uvx) used for CLI invocation
✓ API credentials managed via standard environment variable pattern