ClawHub
Back to skill

Security audit

Seedance 2.5 社交电商回复视频

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent AI Hive video-generation helper, with disclosed API use, user-provided media upload, local API-key storage, and output download behavior aligned to its stated purpose.

Install only if you are comfortable sending the entered question, approved answer, metadata, and selected product media to AI Hive. Use --preview first, redact personal data and business-sensitive details, and prefer the environment variable or the chmod-protected config file for the API key.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill invokes a local Python script that supports authentication, status polling, file inputs, and API-key handling, which implies shell, network, file, and environment access without any declared permission boundary. This is dangerous because an agent may execute the skill with broader capabilities than the user expects, enabling outbound data transfer and local file access during handling of potentially sensitive customer comments, product assets, and credentials.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The description contains very broad 'use this skill for' phrasing across many platforms, use cases, and competitor comparisons, without a clear trigger boundary or narrow activation criteria. In an agent setting, this can cause over-invocation on loosely related requests, increasing the chance of unintended execution, unnecessary data handling, or accidental submission of sensitive commerce and user-content inputs to external services.

Missing User Warnings

Medium
Confidence
82% confidence
Finding
The script uploads user-supplied questions, answers, images, and optional videos to remote AI Hive endpoints, but it does not present a clear user-facing consent/privacy warning at the point of transmission. In a social-commerce context, these inputs can contain personal data or sensitive business content, so silent remote transfer increases the risk of inadvertent privacy disclosure and compliance violations.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.