validation-quarantine

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly processes external "rawArticles" via validateBatch and scores/reads untrusted article fields (title, url, domain — e.g., checking user-generated platforms like blogspot/wordpress/medium), so it ingests and interprets public third‑party/user‑generated content as part of its workflow.