elevenlabs-reference-architecture

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow ingests the outsider-provided text argument passed into TTSService.generate(text, ...) (and generateLongText(...)) from the API layer (e.g., /api/tts), meaning user-submitted free text is directly sent into ElevenLabs text-to-speech at runtime.