Text To Video Kling Ai

• Outcome: Review
• Latest audit: May 29, 2026, 9:33 PM
• Version:
  1.0.0

Overview

The skill is a real cloud video-generation workflow, but it automatically creates third-party sessions and can send prompts or media to NemoVideo with insufficient user-facing consent.

Install only if you are comfortable with prompts, files, generated media, and session metadata being sent to mega-api-prod.nemovideo.ai. Use a dedicated NEMO_TOKEN if possible, avoid private or regulated media, and require the agent to ask before connecting, uploading files, starting renders, or using 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
• 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: 82% confidence
• Finding: The activation examples and phrasing are broad enough that ordinary creative requests could trigger this skill unexpectedly. Because the skill can initiate network access and session creation with credentials or anonymously acquired tokens, overbroad invocation increases the risk of unintended third-party data transmission and backend actions without sufficiently clear user intent.

Vague Triggers

Medium

• Confidence: 93% confidence
• Finding: The fallback rule routing 'Everything else' to the SSE action creates an effectively catch-all activation path. In a skill that can send arbitrary user prompts to a remote editing backend, this broad routing can cause accidental invocation, excessive data sharing, and unintended edits or job creation from unrelated user requests.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill instructs the agent to use an environment token if present or silently obtain an anonymous token from a third-party service, but explicitly tells it to hide the technical details from the user. That combination is dangerous because it authorizes outbound authenticated requests and account/session creation without transparent consent, which can expose user identity, billing context, or usage metadata to an external service.

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The skill supports file upload and cloud rendering but does not prominently warn users that their prompts and media files will be transmitted to a third-party backend for processing. This is especially risky because accepted file types include videos, images, and audio that may contain sensitive or copyrighted material, and the cloud workflow persists session state and render artifacts remotely.

VirusTotal

64/64 vendors flagged this skill as clean.