fittings-pubmed-search

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

• Third-party content exposure detected (low risk: 0.10). In SKILL.md, the required runtime workflow calls the pubmed.search endpoint at https://fittings.sh/v1/pubmed/search with a user-supplied query, but it only sends that query as a structured search term and then returns PubMed citations (trusted first-party NCBI E-utilities outputs), not outsider-authored free text for the agent to ingest.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's docs explicitly show fetching and then executing remote code ("curl -sO https://fittings.sh/skill/call.mjs" followed by "node call.mjs"), so https://fittings.sh/skill/call.mjs is a runtime external dependency that executes fetched code.

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

• Direct money access detected (high risk: 1.00). The skill explicitly describes metered paid calls and the payment flow: it returns a 402 with a signed quote, requires the caller to "sign an authorization", and provides example instructions to use a wallet (USDC on Base) by exporting FITTINGS_PRIVATE_KEY. It also references evm tooling (@x402/evm) and warns the key sits in plaintext. Those are concrete crypto/payment signing capabilities (wallet + private key + signing auth) rather than purely generic tooling, so the skill grants direct financial execution authority.