ClawHub
Back to skill

Security audit

analyze video by qwen

Security checks across malware telemetry and agentic risk

Overview

The skill does what it claims, but it also gives broad local command permissions and sends user-selected videos to a cloud provider without strong disclosure or consent controls.

Review this before installing. Use a dedicated DashScope key, expect provider quota or charges, and only analyze videos you are comfortable sending to Alibaba Cloud DashScope/Qwen. Remove or narrow the .claude settings that allow arbitrary python3 and chmod commands, and pin the dashscope dependency before routine use.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
87% confidence
Finding
The skill documentation indicates capabilities to read local files and access remote URLs, but it does not declare permissions or prominently scope those behaviors. This can mislead users about what data the skill may access and transmit, especially because local video files may contain sensitive content and remote URL handling introduces external network access.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill advertises cloud-based video analysis and remote URL support but does not clearly warn that video content, derived frames, prompts, and possibly fetched remote media will be transmitted to an external provider. Users may unknowingly submit sensitive or regulated data, creating privacy, confidentiality, and compliance risks.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The script sends user-supplied video content to DashScope's remote multimodal API, but it does not present an explicit privacy/data-disclosure warning or require affirmative consent before upload. This is dangerous because videos may contain sensitive personal, proprietary, or regulated information, and users may reasonably assume local-only processing when providing a local file path.

Unpinned Dependencies

Low
Category
Supply Chain
Content
dashscope
Confidence
94% confidence
Finding
dashscope

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.