Editor Pc

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:41 PM UTC
• Version:
  1.0.0

Overview

This is a disclosed cloud video-editing skill that sends chosen media and edit prompts to NemoVideo, with no evidence of hidden code, exfiltration, or destructive behavior.

Install only if you are comfortable sending selected footage, edit instructions, session state, and render jobs to mega-api-prod.nemovideo.ai. Avoid uploading private or regulated video unless you trust NemoVideo's handling of it, use a dedicated token when possible, and ask the agent to confirm uploads, exports, and credit-impacting actions before proceeding.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The example trigger phrases are generic enough that ordinary conversation like asking to edit footage or export a file could activate the skill unexpectedly. In this skill, unintended activation is more sensitive because it can immediately initiate setup against a third-party video-processing service and encourage users to upload media without a clear consent boundary.

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: The routing table contains a broad catch-all condition ('Everything else') that can funnel many unrelated prompts into the SSE editing workflow. Because this workflow can send user text to a remote backend and alter session state, ambiguous routing increases the risk of unintended data disclosure and accidental remote actions.

Missing User Warnings

Medium

• Confidence: 98% confidence
• Finding: The skill description markets easy video editing but does not clearly warn users that uploaded videos and prompts are sent to a remote third-party processing API. This omission is significant because raw video footage may contain sensitive personal, workplace, or biometric information, and users may reasonably assume local or first-party handling from the description alone.

Natural-Language Policy Violations

Low

• Confidence: 77% confidence
• Finding: Forcing GUI translation behavior without user opt-in can misrepresent backend output, especially across languages, and may cause the agent to take actions based on translated intent rather than the user's exact wording. In an editing/export workflow, this is lower severity than direct code execution, but it can still lead to mistaken operations or misleading user communication.

VirusTotal

66/66 vendors flagged this skill as clean.