orpheus-tts-voice-trainer

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s workflow ingests outsider-authored free text at runtime via brave-search and get-subtitles/watch/ingest-movie in steps that “prefer authentic sources first” to find movie/audiobook/dialogue evidence, meaning it can read arbitrary external transcripts/snippets into the agent.