video-audio

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

• Third-party content exposure detected (medium risk: 0.30). scripts/generate_audio.py reads LLM-ingestible free text from the user-provided storyboard_dir/scenes.json narration fields at runtime (scene["narration"] is sent to ElevenLabs TTS via generate_tts), so an outsider who can submit/alter scenes.json can inject arbitrary narration text.