icd10-lookup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill's examples instruct the user to place a private key directly into a shell command/env var (export FITTINGS_PRIVATE_KEY=0x...), which requires embedding secret values verbatim in commands and leaves them in plaintext where an LLM or logs could exfiltrate them.

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

• Potentially malicious external URL detected (high risk: 0.90). The SKILL instructs fetching and then executing remote code with curl -sO https://fittings.sh/skill/call.mjs (and the skill warns to re-install from that exact URL before spending), which is a runtime remote-code fetch-and-execute dependency.

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

• Direct money access detected (high risk: 1.00). The skill includes an explicit crypto payment flow: it requires/accepts a private key (FITTINGS_PRIVATE_KEY), instructs using a wallet holding USDC on Base, and has the client sign an authorization to settle quotes (i.e., perform on-chain/off-chain payment actions). Those concrete instructions and keys enable the agent to sign/pay funds, so this grants direct financial execution capability.