cro-methodology

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly requires gathering and interpreting untrusted, user-generated third-party content—e.g., "Gather market intelligence from competitors, reviews, and other industries" in "The CRO Process" and "Secondary Sources" (reviews, social media, Reddit, Twitter) in references/OBJECTIONS.md and references/RESEARCH.md—which the agent is expected to read and use to form hypotheses and drive A/B tests.