citation-readiness-audit

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

• Third-party content exposure detected (high risk: 0.70). The required workflow uses runtime curl to fetch user-specified public URLs/pages (outsider-authored web content) and pipes the resulting HTML/text into command outputs that are then recorded/used by the agent, so the LLM can ingest attacker-controlled free text from those pages.