hts-lookup

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.80). These URLs point to an externally hosted executable script (https://fittings.sh/skill/call.mjs) and the docs instruct downloading/running it and placing a private key in plaintext — downloading/running remote scripts and exposing keys are high-risk behaviors even if the domain appears legitimate.

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

• Third-party content exposure detected (medium risk: 0.30). Runtime path hts.lookup→ call.mjs builds a request from user-supplied code and then ingests the fittings.sh response (JSON text from https://fittings.sh/v1/hts/lookup) without selecting any specific pre-defined item first.

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

• Potentially malicious external URL detected (high risk: 0.90). The SKILL.md explicitly instructs fetching and then running remote code ("curl -sO https://fittings.sh/skill/call.mjs" and executing node call.mjs), so https://fittings.sh/skill/call.mjs would supply code executed at runtime.

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

• Direct money access detected (high risk: 1.00). The skill explicitly describes a payment flow that requires signing on-chain authorizations and using a wallet: it instructs setting FITTINGS_PRIVATE_KEY, states the caller needs a wallet holding USDC on Base, describes signing an authorization and a facilitator broadcasting/paying gas, and references EVM-related packages (@x402/evm). These are specific crypto/blockchain payment capabilities (wallet signing/spending), i.e. direct financial execution authority.