transcription

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes examples that pass tokens verbatim (e.g., --hf_token YOUR_HF_TOKEN and use_auth_token='YOUR_HF_TOKEN'), which instructs embedding API secrets directly into commands/code and thus requires the LLM to handle/output secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.70). This skill explicitly accepts remote, user-generated URLs (e.g., YouTube) as runtime inputs (see SKILL.md "Supports local files, YouTube URLs", commands/transcription.md and templates/implementation-template.md which validate/run transcribe-anything on arbitrary URLs and show reading the resulting .txt transcript), so the agent ingests untrusted third‑party audio/text that could contain instructions and materially influence subsequent agent actions.