scribe

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly ingests untrusted user-generated media and external storage URIs for transcription (see examples/fast-mode-node.md for URL/file uploads and examples/batch-webhook-pipeline.md for S3 input URIs and webhook ingestion), so transcripts derived from third-party content could contain instructions that materially influence downstream actions.