market-brand

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly instructs the agent to pull and analyze raw text from public third‑party pages (e.g., "Homepage", "About page", "Blog (3–5 recent posts)", "Social profiles") using scripts/analyze_page.py and requires quoting that source material to drive scoring and generated copy, so untrusted external content can directly influence agent decisions.