ClawHub
Back to skill

Security audit

电商主图

Security checks across malware telemetry and agentic risk

Overview

This skill sends user-selected product images and prompts to AI Hive for image generation, with disclosed local API-key setup and no hidden or destructive behavior found.

Install only if you are comfortable sending selected product images and prompts to AI Hive, and treat the saved API key like a credential. Review the generated images manually for SKU accuracy and platform-policy compliance before using them commercially.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill advertises shell execution, file access, environment access, and network usage without declaring permissions, which weakens user and platform visibility into what the skill can do. In this context, the commands install packages, read local images, write configuration/API key material, and make authenticated outbound requests, so undeclared capabilities materially increase the risk of unauthorized data access or unintended side effects.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding
The documented purpose is narrowly framed as generating compliant ecommerce main images from SKU truth data, but the described behavior is broader: it can guide API key setup, upload arbitrary images, query arbitrary tasks, and act as a generic prompt/model forwarding interface. That mismatch is dangerous because users may trust the skill with sensitive product assets or credentials under a constrained-use expectation, while the implementation appears to support more general remote operations without the promised SKU-specific validation or policy safeguards.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.