agent-speak

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

• Third-party content exposure detected (high risk: 0.80). The skill accepts arbitrary external media URLs for voice creation (see references/voice/kling.md: the example rawgenai kling voice create "MyVoice" --audio "https://example.com/audio.mp3"), which means the agent can fetch and ingest untrusted third-party content (audio) that could influence downstream behavior (voice cloning/outputs).