ai-multimodal

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly ingests and analyzes open/public third‑party content — e.g., references/video-analysis.md shows direct YouTube URL analysis (types.Part.from_uri with https://www.youtube.com/...) and references/vision-understanding.md shows public image URL ingestion — meaning the agent will read and interpret untrusted, user‑generated web content as part of its workflow.