ClawHub
Back to skill

Security audit

GPT Image 2 图片编辑

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed AI Hive image-editing integration that uploads chosen images, stores an API key only when the user runs setup, and downloads generated results.

Before installing, expect images you select to be uploaded to AI Hive and generated results to be downloaded locally. Use a dedicated AI Hive API key if possible, review the provider account and billing implications, and avoid uploading private or regulated images unless that provider is acceptable for your use case.

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
93% confidence
Finding
The skill invokes a Python script with capabilities to access environment variables, read/write local files, use the network, and execute shell commands, yet no permissions are declared to the user or platform. This creates a transparency and least-privilege problem: operators may approve an image-editing skill without realizing it can also persist configuration, access secrets, and communicate externally.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
The documented purpose is narrowly framed as GPT Image 2 image editing, but the underlying toolchain reportedly supports much broader actions including generic chat, video generation, user/account queries, model enumeration, media upload, and interactive API key setup with local storage. That mismatch is dangerous because users may invoke the skill under false assumptions, while the code can collect credentials, access account metadata, and perform unrelated networked operations outside the declared scope.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.