Video Understanding

• Outcome: Pass
• Latest audit: May 28, 2026, 1:18 PM
• Version:
  1.1.0

Overview

This skill does what it claims: it analyzes user-provided videos with Gemini, with the main consideration being that video URLs or downloaded video content are sent to Google for processing.

Install only if you are comfortable sending the video URL or video content to Google Gemini using your GEMINI_API_KEY. Avoid using it with confidential, private, regulated, or access-restricted videos unless that third-party processing is acceptable for your situation.

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 Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 91% confidence
• Finding: The skill advertises and relies on shell execution, environment access, and file read/write behavior, but it does not explicitly declare permissions. That creates a transparency and policy-enforcement gap: users or orchestrators may invoke the skill without understanding that it can download media, access API keys, and write files to disk.

Vague Triggers

Medium

• Confidence: 82% confidence
• Finding: The trigger language is broad enough to match many ordinary requests involving watching, summarizing, or answering questions about videos. Over-broad invocation can cause the skill to activate unexpectedly, leading to unreviewed downloading of third-party content or transmission of user-provided URLs to external tools and services.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The documentation states in 'How It Works' that non-YouTube URLs are downloaded and uploaded to Gemini File API, but the skill description and trigger text do not clearly disclose this external transfer up front. This is dangerous because users may provide private, authenticated, or sensitive video links without realizing the content will be copied locally and sent to Google's service, creating privacy, compliance, and data-handling risks.

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The skill uploads downloaded video content to Gemini for analysis without any explicit consent flow, warning, or data-handling disclosure. Videos may contain sensitive personal, confidential, or regulated content, so silent transmission to a third-party AI service creates a real privacy and compliance risk.

VirusTotal

58/58 vendors flagged this skill as clean.