battle-card-builder

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

• Third-party content exposure detected (high risk: 0.85). The required workflow explicitly performs OSINT/SIGINT “search” for public evidence (e.g., pricing pages, release notes, reviews, comparison content) and then incorporates that evidence (with URLs/dates) into the LLM-generated battle card, so outsider-authored free text from fetched web sources can enter the LLM context via the investigation/search path.