classify-media-relevance

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

• Third-party content exposure detected (high risk: 0.90). This skill runs "during the parse pipeline, after the AI extracts release content from a fetched page" and "during crawl-mode fetches, when the extractor reads full-page markdown from a linked article," so it ingests alt text and surrounding content from arbitrary public web pages (untrusted third-party sources) and uses that content to decide keep/drop actions that materially affect downstream behavior.