company-radar

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

• Third-party content exposure detected (high risk: 0.85). Outsider free text is ingested via runtime web research (e.g., Tavily search results and scraped pages from arbitrary company-related URLs) and then used to generate the AI executive briefing and report text, which can include untrusted third-party content (public web/Reddit/HN/Twitter/Product Hunt posts) authored by others.