bounce-handling

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly ingests and normalizes webhook payloads from third-party email providers (see "Provider bounce webhook formats" for Amazon SES, Postmark, Resend, SendGrid) and uses those untrusted external payloads to make suppression/retry/auto-pause decisions, so attacker-controlled webhook content could influence agent behavior.