money-discover

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly requires searching and mining user-generated public sources (Product Hunt, Hacker News, X/Twitter, Reddit, forums, review sites) and fetching competitor pages/reviews/URLs as part of its required Phase 2 and Phase 5 workflows, which the agent must read and use to drive decisions—exposing it to untrusted third‑party content that could carry indirect prompt injections.