customer-research

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

• Third-party content exposure detected (high risk: 0.85). The skill’s Mode 2 (“Go Find Research”) workflow instructs the agent to gather and extract verbatim quotes from outsider-authored public web content at runtime (e.g., Reddit/G2/Capterra/forum posts and comments), which the agent would then ingest into the LLM context as readable text.