market-researcher

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

• Third-party content exposure detected (high risk: 0.90). The skill's required workflow (Step 3: "Secondary research") explicitly directs the agent to use public, user-generated sources—e.g., Reddit, Twitter, App Store reviews, G2, Capterra, Trustpilot, competitor websites and job postings—which the agent is expected to read and interpret to drive competitor analysis and recommendations, creating a clear avenue for indirect prompt injection.