video-digest-ingest

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

• Third-party content exposure detected (medium risk: 0.30). The required workflow ingests outsider-authored YouTube video metadata and transcripts via yt-dlp and Whisper from user-supplied URLs.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (medium risk: 0.30). The skill executes uvx --from mlx-whisper mlx_whisper, which dynamically fetches and runs a package from PyPI via a package manager without a version lock or repository locator, falling outside the strong/official provenance boundary.