Video Script

Security checks across malware telemetry and agentic risk

Overview

This instruction-only skill is broadly coherent with a cloud video/script generation service, but users should understand it sends prompts and uploaded files to an external NemoVideo backend and may use service credits.

Install only if you are comfortable sending prompts and uploaded TXT, DOCX, PDF, MP4, or other media to mega-api-prod.nemovideo.ai for cloud processing. Avoid sensitive documents unless you trust the service’s privacy and retention practices, and use a limited or anonymous token where possible because exports and rendering may consume credits.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (5)

Description-Behavior Mismatch

High
Confidence
86% confidence
Finding
Advertising a script-generation skill as producing '1080p MP4 files' obscures that the skill actually supports broader media rendering and export operations. This kind of capability overreach is dangerous because it masks the real operational scope and can cause users to provide content under false assumptions about what the skill does.

Description-Behavior Mismatch

Medium
Confidence
86% confidence
Finding
Advertising a script-generation skill as producing '1080p MP4 files' obscures that the skill actually supports broader media rendering and export operations. This kind of capability overreach is dangerous because it masks the real operational scope and can cause users to provide content under false assumptions about what the skill does.

Context-Inappropriate Capability

Medium
Confidence
89% confidence
Finding
The skill is instructed to silently obtain anonymous tokens and manage credit-backed backend access before servicing requests, which is not justified by a narrow 'script generation' claim. This introduces undisclosed account/session creation and third-party service interaction, creating consent, billing, and privacy risks if the agent acts on behalf of the user without explicit approval.

Vague Triggers

Medium
Confidence
80% confidence
Finding
The trigger phrases are broad enough to match ordinary conversation, making accidental activation more likely. In this skill, accidental activation is more dangerous because activation leads directly into remote connection setup and possible upload/render workflows, not just a harmless local text transformation.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill does not clearly warn users that their prompts and uploaded files are sent to a cloud backend, despite instructing the agent to establish remote sessions and upload content. This is a meaningful privacy and consent issue, especially because users may share sensitive documents believing they are only asking for local script assistance.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal