video-steal

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

• Third-party content exposure detected (low risk: 0.10). The workflow downloads and processes external video reels from URLs (YouTube, Instagram, TikTok) provided by the user or an external catalog search, which can contain untrusted audio and visual content that is analyzed via ffmpeg and inspected by the agent.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to query external URLs (https://api.elevenlabs.io/...) using cURL for audio isolation and sound generation, which constitutes a weak/unknown/suspicious operational dependency because ownership/distribution of the API endpoint is assumed dynamically via user-provided credentials without a verified first-party ecosystem constraint.