analyze-account-signals

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow explicitly prioritizes reading primary evidence from CRM/Knowledge/Files/Email/Meeting Transcripts/Internal Messaging/Calendar, and those sources can include outsider-authored free text (e.g., customer emails or meeting transcripts) that would be ingested into the agent’s LLM context during evidence collection.