video-localization

Warn

Audited by Snyk on Feb 20, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 1.00). The SKILL.md Quick Start and multiple example requests include a "video_urls" field (e.g., https://example.com/my-video.mp4) that the each::sense /chat endpoint fetches and processes, meaning the skill ingests arbitrary external (potentially untrusted) video/audio content which the agent transcribes and uses to drive dubbing, voice cloning, subtitles, and workflow actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 20, 2026, 07:58 AM