brief-recruitment-profile

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs fetching and summarizing content from public third‑party sources — e.g., company careers pages, LinkedIn (including Wayback snapshots), Glassdoor "Interviews" (user reviews), Indeed/AngelList/Wellfound listings and optional Reddit/HN via the social-feedback companion — and then uses that untrusted, user‑generated content to drive decisions like flagging ghost roles and determining hiring/headcount signals.