Youtube Video Editor Transitions

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

Overview

This skill is a disclosed cloud video-editing connector, but users should know their media and prompts are sent to NemoVideo servers.

Install only if you are comfortable sending uploaded video, audio, images, edit prompts, and related metadata to NemoVideo's cloud service. Avoid confidential or unreleased media unless you trust that service's privacy and retention practices, and confirm before ambiguous edits or exports that may consume credits.

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 (4)

Description-Behavior Mismatch

Medium

• Confidence: 95% confidence
• Finding: The skill is presented as a narrow 'add transitions and export' tool, but the documented capabilities are materially broader, including general editing, text overlays, audio/BGM handling, previews, and iterative timeline manipulation. This scope mismatch can mislead users and calling agents into granting access or invoking the skill in contexts beyond what the manifest suggests, increasing the chance of unintended data handling and over-broad activation.

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: Routing 'Everything else' to the SSE editing action creates an overly broad trigger surface, so unrelated or ambiguous user prompts may be sent to a remote editing backend. That can cause accidental activation, unexpected transmission of user content, and actions outside the user's intent, especially because the backend accepts free-form editing instructions.

Vague Triggers

Medium

• Confidence: 86% confidence
• Finding: The invocation text says users can 'just describe what you're after,' which is vague and does not clearly bound when the skill should be used. In an agent ecosystem, ambiguous activation language increases the risk that ordinary conversational requests are incorrectly routed into this skill, potentially sending prompts and media to an external service without sufficiently specific user intent.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: The skill sends uploaded media and editing instructions to a remote cloud processing backend, but the user-facing setup and workflow text does not provide a prominent upfront warning at the point of use. This is a privacy and consent issue: users may believe the tool operates locally or may not realize their videos, metadata, and prompts are transmitted to third-party infrastructure.

VirusTotal

66/66 vendors flagged this skill as clean.