integrate-midtrans-payments

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires fetching live Midtrans documentation at runtime (starting with https://docs.midtrans.com/llms.txt and the linked docs.midtrans.com pages) to drive its prompt/instruction decisions, so it depends on external content that directly controls agent behavior.

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

• Direct money access detected (high risk: 1.00). The skill is explicitly and specifically about integrating with a payment gateway (Midtrans). It prescribes using Midtrans products (Snap, Core API, BI-SNAP, Payment Link, Subscriptions), creating payments, using server keys/Basic Auth, managing access tokens and notification signatures, handling callbacks/webhooks, issuing refunds (with refund_key/partnerRefundNo), and performing sandbox/live smoke tests that replay webhooks and execute provider API calls. Those capabilities are concrete, payment-provider-specific operations (creating/refunding/validating payments and managing provider credentials) rather than generic tooling. Under the Core Rule this is a Direct Financial Execution capability (payment gateway integration).