agent-checkouts

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs the agent to ask the user for a server-side API key and then run a command embedding that key verbatim (e.g., node scripts/checkout.mjs set-key sk_production_XXXXXXXX), which requires the LLM to handle/output the secret value directly.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill makes runtime calls to the Crossmint Agent Checkouts API (https://www.crossmint.com/api/unstable/agent-checkouts and https://staging.crossmint.com/api/unstable/agent-checkouts), whose JSON responses (notably pendingUserAction.responseSchema) directly drive what prompts/values the agent must submit, and the skill requires those endpoints to operate.

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 execute real purchases via Crossmint's Agent Checkouts API. It requires and stores a server-side API key, creates checkouts, and provides a checkout-respond command that, when submitted, immediately charges payment and places a real order. This is a purpose-built payment/checkout integration (not a generic browser or HTTP tool) that grants direct financial execution authority.