gather-business-context

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s workflow explicitly performs runtime retrieval/extraction from “enabled or provided apps” (via the required data-analytics:user-context preflight and subsequent connector searches), which can ingest outsider-authored free text such as public web content, tickets/issues, or other third-party messages into the LLM context.