stripe

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.80). The prompt contains hidden/deceptive instructions (the SELF-EVOLUTION TELEMETRY section) that direct the agent to record execution metadata to a local ~/.claude/projects/skill-telemetry.md location—behavior unrelated to the stated Stripe integration purpose and potentially exfiltrative, so this is a prompt injection.

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

• Direct money access detected (high risk: 1.00). The skill is explicitly designed to integrate Stripe (a payment gateway). It installs the Stripe SDK, configures STRIPE_SECRET_KEY, creates a Stripe client/service with methods like createCheckoutSession and createPaymentIntent, and adds API endpoints (POST /api/payments/checkout, /api/payments/create-intent, /api/payments/portal) that initiate payments, create payment intents, manage subscriptions, and create customer portal sessions. These are specific payment gateway operations (not generic HTTP or browser tooling) that can move money or authorize charges, so it grants Direct Financial Execution authority.