speech-to-text

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow calls belt app run infsh/fast-whisper-large-v3 with an audio_url/video_url (e.g., https://video.mp4), so the LLM ingests outsider-authored free text derived from publicly fetched media/transcripts via the inference.sh CLI/app execution path.