customer-research

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

• Third-party content exposure detected (high risk: 0.85). This skill’s required workflow explicitly supports “Mode 2: Go Find Research,” which instructs the agent to read online customer content (e.g., Reddit/G2/forums/review sites) and extract verbatim quotes, so at runtime outsider-authored free text from those sources can be ingested into the LLM context.