ClawHub
Back to skill

Security audit

Image Generation Agent

Security checks across malware telemetry and agentic risk

Overview

This is a coherent remote image-generation skill with disclosed AgentPMT use and no evidence of hidden local execution or malicious behavior.

Install only if you are comfortable sending prompts, reference images, and public image URLs to AgentPMT and its image provider. Do not submit confidential, regulated, or private images unless your account terms and retention controls fit your needs, and use short expiration periods for generated files when possible.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill states that generated images are auto-saved and exposed through signed download URLs, but it does not present a prominent upfront warning that user-supplied reference images and outputs are transmitted to and stored by a remote third-party service. This can cause users or calling agents to send sensitive images under the mistaken assumption that processing is local or ephemeral, leading to privacy, confidentiality, or compliance exposure.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The schema explicitly allows `reference_images` with `source_kind: "url"` and a public HTTPS `url`, but it does not warn users that the backend may fetch remote content. That can expose user-provided URLs and associated metadata to external services, create privacy surprises, and enable SSRF-style risk if URL fetching is not tightly restricted by the implementation.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.