cyber-audit

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

• Third-party content exposure detected (high risk: 0.70). The required workflow instructs fetching/using the advisory’s full text (“fetch the CVE/advisory details… or use the full text the user provided”), which at runtime can include outsider-authored free text (e.g., public advisory content from authoritative sources) that is then used to ground the audit and generate the LLM’s context.