video-editing

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

• Third-party content exposure detected (medium risk: 0.30). The required runtime workflow ingests user-provided video/audio transcript content via the “Layer 2: Organization (Gemini / Codex) — Transcribe and label” step, which means outsider-authored free text (the transcript or user input used to prompt it) can reach the LLM without selecting a trusted item first.