elevenlabs-music

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s required runtime workflow ingests user-supplied JSON input fields (notably prompt and optionally text for voiceover) to generate music/voice via belt app run elevenlabs/music and belt app run elevenlabs/tts, so the LLM reads outsider-authored free text provided at execution time.