purview-communication-compliance

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

• Third-party content exposure detected (high risk: 0.85). The skill is about supervising and ingesting message content from Microsoft 365/connected third-party communications (e.g., Exchange/Teams/Viva Engage and connectors like Slack/Bloomberg), so at runtime it can feed outsider-authored free text (other users’ messages/comments) into the agent/LLM context for classification/triage.