gemini-omni-flash-api

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

• Third-party content exposure detected (medium risk: 0.30). The runtime path scripts/video/generate_video.py ingests outsider-provided free text as the prompt (from CLI args.prompt, --prompts-file, or --batch JSON) and sends it to the model via client.interactions.create(input=[...,"type":"text","text": prompt ...]) without requiring the user to pre-select a specific trusted item.