spark-persona-support-agent

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs the agent to list and read shared inbox items and full email threads (e.g., "spark emails shared@company.com:Inbox" and "spark thread "), which are untrusted, user-generated third-party content that the agent must interpret to decide actions and drafts, enabling indirect prompt injection.