video-from-reference

Warn

Audited by Snyk on Feb 21, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md requires providing image_urls (reference images) which the agent fetches/ingests from arbitrary external URLs and those untrusted images are directly used to influence the model's output, creating a vector for indirect prompt injection.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 21, 2026, 08:22 AM