skill-creator 安全扫描报告 — 低风险 | ClawSafe

5 /100

skill-creator

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

This is a legitimate skill-creation meta-skill providing guidance and utilities for creating AI skills. The scripts perform standard operations (directory creation, YAML validation, zip packaging) using only Python standard library with no security concerns.

技能名称skill-creator

分析耗时32.7s

引擎pi

✓

可以安装

Skill is safe to use. Consider adding an `allowed-tools` declaration in SKILL.md frontmatter to document expected filesystem access patterns for the bundled scripts.

安全发现 1 项

严重性	安全发现	位置
低危	Missing allowed-tools declaration 文档欺骗 The SKILL.md frontmatter lacks an allowed-tools field to document the filesystem access required by bundled scripts. While not a security risk, declaring this would improve transparency. `---` → Add allowed-tools field if specific tool permissions are required	`SKILL.md:1`

资源类型	声明权限	推断权限	状态	证据
文件系统	`NONE`	`WRITE`	✓ 一致	scripts/init_skill.py:96 - mkdir(parents=True) and write_text()
文件系统	`NONE`	`READ`	✓ 一致	scripts/package_skill.py:28 - rglob('*') for zip creation

2 项发现

🔗

中危外部 URL 外部 URL

http://www.apache.org/licenses/

LICENSE.txt:4

🔗

中危外部 URL 外部 URL

http://www.apache.org/licenses/LICENSE-2.0

LICENSE.txt:196

目录结构

8 文件 · 48.5 KB · 1178 行

Python 3f · 507L Markdown 3f · 465L Text 1f · 201L JSON 1f · 5L

├─ ▾ 📁 references

│ ├─ 📝 output-patterns.md Markdown 82L · 1.8 KB

│ └─ 📝 workflows.md Markdown 27L · 818 B

├─ ▾ 📁 scripts

│ ├─ 🐍 init_skill.py Python 303L · 10.6 KB

│ ├─ 🐍 package_skill.py Python 110L · 3.2 KB

│ └─ 🐍 quick_validate.py Python 94L · 3.4 KB

├─ 📋 _meta.json JSON 5L · 132 B

├─ 📄 LICENSE.txt Text 201L · 11.1 KB

└─ 📝 SKILL.md Markdown 356L · 17.4 KB

安全亮点

✓ Uses only Python standard library (no external dependencies)

✓ No credential theft, API key harvesting, or environment variable access

✓ No base64 encoding, obfuscation, or eval() patterns

✓ No network requests or data exfiltration

✓ No shell execution (subprocess/os.system)

✓ No sensitive path access (~/.ssh, ~/.aws, .env)

✓ No remote script execution (curl|bash, wget|sh)

✓ No persistence mechanisms or backdoors

✓ YAML validation uses safe yaml.safe_load()