sk-competitors

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly requires WebSearch and directs agents in Phase 2 (Wave 2: "Review Mining" and "Forum & Community Mining") and Wave 1/3 to scrape and mine untrusted user-generated sources (G2, Capterra, Reddit, Hacker News, Quora, Twitter/X, Product Hunt, app stores, arbitrary competitor websites) and then uses those findings to drive synthesis, battle cards, pricing, and strategic actions—so third‑party content is ingested and can materially influence agent decisions.