youtube-api

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill includes literal API-key-like strings (e.g., export FETCHER_API_KEY="bby_live_xxxxxxxxxxxx" and an MCP header with "Authorization: Bearer bby_live_...") and shows embedding them into Authorization headers, which requires handling or echoing secrets verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime ingests outsider-authored free text via query parameters like query passed into YouTube search endpoints (e.g., /api/search/video with query=...) which can be supplied by the end user, and the fetched results/comments may contain further free text.

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

• Direct money access detected (high risk: 1.00). The skill explicitly includes payment/top-up endpoints and on-chain payment mechanics: it references POST /api/credits/topup and a topup_credits function, describes an x402 pay-per-call flow that requires USDC on specific blockchains (Base, Polygon, Arbitrum, Monad, Solana), and notes an @x402/fetch client that "signs and retries automatically" (implying transaction signing). These are explicit crypto/payment execution capabilities rather than generic data-only operations.