email-campaign

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

• Third-party content exposure detected (medium risk: 0.30). In the email-campaign skill’s required workflow, outsider-authored free text (the user’s CLI-provided fields like --domain/--email/--firstName/--lastName/--company) is sent into multiple POST request bodies to third-party merchants (Fiber/Hunter/Apollo/Abstract Company Enrichment), so the workflow can ingest arbitrary user-submitted strings without first selecting a trusted item.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly executes payments: each step is paid "per-call through MPP merchants and via the SELAT Router" and the CLI compiles steps into selat-pay calls that settle through the SELAT Router (funded Circle Gateway balance). The docs include concrete selat-pay POST commands, per-step prices, and settlement details — this is an explicit payment execution capability (moving money), not a generic HTTP or automation tool.