mailchimp

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime calls to https://gateway.maton.ai/mailchimp/... (e.g., lists/members/campaign content) ingest Mailchimp-provided JSON fields like html/plain_text and error detail/message strings into the agent context; those strings are outsider-authored content from an external service (Mailchimp) rather than the operating user.