ClawHub
Back to skill

Security audit

Mj Gallery

Security checks across malware telemetry and agentic risk

Overview

This skill openly generates images, saves them locally, and publishes a gallery, so the main risk is privacy exposure from what the user chooses to generate.

Install only if you are comfortable with generated images, prompt summaries, filenames, and task metadata being saved in the workspace and published through the gallery link. Avoid confidential prompts, private reference images, client work, or copyrighted material unless you control the deployment destination and can remove the published gallery. Use a scoped or revocable MX_AI_API_KEY.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
95% confidence
Finding
The skill description explicitly promises automatic local download, archival, and public gallery deployment, but it does not require an explicit user consent step or warning before publishing generated content. This creates a privacy and data exposure risk because user prompts and generated images may be made publicly accessible by default, potentially exposing sensitive or copyrighted material.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The mandatory workflow requires deploying `/workspace/mj_gallery/` to a public URL and returning that link, yet the instructions do not include a privacy or integrity notice before publication. In context, the gallery also embeds prompt-derived text and task metadata in `index.html`, increasing the risk of unintended disclosure beyond the image itself.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.