buy-agent-commerce-guard

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

• Third-party content exposure detected (high risk: 0.85). The required workflow explicitly fetches multiple externally hosted free-text/JSON resources at runtime from https://agent-commerce-guard.vercel.app/* (e.g., package-metadata.json/status.json/pay.json/base-mcp.json), and then ingests/summarizes that content into the agent’s LLM context, creating an indirect prompt-injection risk from outsider-authored web content.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches runtime JSON/APIs (e.g., https://agent-commerce-guard.vercel.app/base-mcp.json) to determine payment requirements and approval flows, so external content directly controls agent decisions at runtime.

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 initiate and verify on-chain payments: it defines exact payment constants (network, chain ID, USDC token contract, receiver address, and exact amount), prescribes payment routes (wallet, manual transfer, MCP/x402 client), requires showing a compact payment confirmation, and instructs verifying transaction hashes via payment-status and download endpoints. These are specific payment execution/verification operations (on-chain USDC transfer), not generic tooling, so it grants direct financial execution capability.