ai-avatar-video

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

• Third-party content exposure detected (high risk: 0.80). SKILL.md explicitly shows workflows that run infsh apps with user-supplied image_url/audio_url/video_url (e.g., infsh/fast-whisper-large-v3 transcription and subsequent TTS/lipsync steps), which fetch and process arbitrary public media, so untrusted third-party content could influence processing and behavior.