stinkyboy

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

• Third-party content exposure detected (low risk: 0.10). The agent integrates with an MCP catalog tool service where third-party tools can be searched and executed via dynamic synthesis, but it does not natively ingest unprompted external feeds or queues.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the MCP client to connect to a remote endpoint at https://mcp.stinkyboy.co/mcp to load operational tools and instructions at runtime, which represents an unverifiable external dependency with unknown third-party publisher provenance.

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

• Direct money access detected (high risk: 1.00). The skill integrates directly with crypto wallets and payment protocols (USDC on Base and Solana via x402 v2) to execute payments per tool call ("paid per call in USDC via x402", "Schemes: exact, upto, batch-settlement, auth-capture"), which falls under the category of crypto/blockchain wallets and payment execution.