Video Editor Transitions

• 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 selected media and edit prompts to Nemo Video for processing, with no local executable code or hidden install behavior found.

Install this only if you are comfortable sending your video files, URLs, edit prompts, and generated project state to Nemo Video's cloud service. Use a dedicated Nemo token where possible, avoid uploading sensitive or third-party footage without permission, and confirm ambiguous requests before letting the skill process or export media.

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
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Description-Behavior Mismatch

Medium

• Confidence: 92% confidence
• Finding: The skill is presented as a narrow transition-editing tool, but the instructions expose substantially broader capabilities including remote media ingestion by URL, audio/text editing, state inspection, credits queries, and generic timeline manipulation. That mismatch can mislead users and host systems about what data and actions the skill may perform, increasing the chance of over-broad invocation and unexpected transmission of user content to the backend.

Vague Triggers

Medium

• Confidence: 88% confidence
• Finding: The catch-all rule routes 'everything else' to the SSE editing path, which can cause loosely related or ambiguous user messages to trigger backend editing actions. In a skill that uploads media and sends prompts to a remote service, broad routing increases the risk of unintended processing, privacy surprises, and accidental modification/export workflows.

Vague Triggers

Medium

• Confidence: 80% confidence
• Finding: Phrases like 'Share your video clips and I'll get started' and broad example triggers encourage activation without clearly bounding what will happen next. Overly permissive invocation language can cause users to initiate backend connections or uploads before understanding that remote processing and session creation will occur.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: Although the document later mentions server-side rendering, the skill description does not clearly warn up front that uploaded videos, prompts, and possibly URLs are sent to a remote backend service. This is a privacy and transparency issue because users may provide sensitive media under the impression that editing is local or more limited than it is.

Natural-Language Policy Violations

Medium

• Confidence: 84% confidence
• Finding: Creating sessions with a hard-coded language of 'en' without user choice can lead to misprocessing of non-English requests and unintended transmission of content under incorrect localization assumptions. While not a direct code execution risk, it can degrade correctness, user consent, and safe handling for multilingual prompts.

VirusTotal

66/66 vendors flagged this skill as clean.