40Medium
This skill is vulnerable to prompt
Behavior Analysis
Description verified — behavior matches claim
Claims to do
📸 QQ空间相册管理器: > QQ空间相册的自动化管理工具，支持扫码登录、相册浏览、照片上传/下载、相册创建等功能。
Actually does
This skill executes a Python script (`qzone_photos.py`) to manage QQ Zone photo albums. It uses `python3` to perform actions like logging in via QR code, listing/browsing albums, uploading/downloading photos, and creating albums. It requires and uses a `cookies.json` file for authentication, which contains QQ Zone session tokens (`p_skey`, `skey`, `uin`). It interacts with QQ Zone's non-official APIs.
Skill Info
Namewscats/qq-zone-photo
Registryclawhub
Version1.0.3
PURLpkg:clawhub/wscats/qq-zone-photo@1.0.3
Stars3
Downloads11,417
Source
SHA-256503180dda8a1...
SHA-17c931ecdb51b...
MD52f09a706fec4...
TLSHe471746aac3a...
Size3,481 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
ClawHubDocs
openclaw skills install wscats/qq-zone-photoAssessments (2)
AI Agent Traps ↗Prompt Injection Detectedmediumprompt injection
DeBERTa classifier detected prompt injection (confidence: 0.75)
75% confidence
Potential for SSRF, Path Traversal via Unsanitized Parametersmediumcommand execution
The skill executes a Python script with parameters like '--url', '--output', and '--photo'. If the underlying 'qzone_photos.py' script does not properly sanitize these inputs, a malicious AI agent could exploit them for Server-Side Request Forgery (SSRF) via '--url', arbitrary file writes via '--output', or arbitrary file reads via '--photo'.
70% confidenceline 55
python3 scripts/qzone_photos.py --action download --url "URL" --cookies cookies.json --output "PATH" python3 scripts/qzone_photos.py --action upload --photo "/path/to/image.jpg" --album-id "ALBUM_ID" --cookies cookies.json
Handling of Sensitive QQ Zone Authentication Cookieslowcredential theft
The skill explicitly handles QQ Zone authentication cookies, which grant full account access. While the skill states credentials are stored locally and not uploaded, the presence and management of such sensitive data (e.g., via the 'login' action) makes it a potential target for misuse by a malicious AI agent or other system vulnerabilities.
80% confidenceline 4
metadata: {"openclaw":{"emoji":"📸","requires":{"auth":["qq-cookies"]}}}
Cookie 文件包含 QQ空间完整访问权限，请勿泄露或分享Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/clawhub/wscats/qq-zone-photo)HTML
<a href="https://mondoo.com/ai-agent-security/skills/clawhub/wscats/qq-zone-photo"><img src="https://mondoo.com/ai-agent-security/api/badge/clawhub/wscats/qq-zone-photo.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/clawhub/wscats/qq-zone-photo.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow