batch-research

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

• Third-party content exposure detected (high risk: 0.90). The skill generates and dispatches open-web URLs (e.g., via generateHNUrls and explicit researcher calls to sites like https://news.ycombinator.com, https://drafts.miantiao.me/, https://www.solidot.org/, https://anthropic.com/, etc.), causing the researcher agent to fetch and read untrusted, user-generated third-party content as part of its workflow.