handle-complaint

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly ingests and reads customer-generated content from Gmail threads and HubSpot tickets (and searches HubSpot/PayPal for related records) as part of its workflow to draft responses and recommend actions, exposing the agent to untrusted third-party content that could contain malicious instructions.