crypto-bd-agent

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow includes “Web Scraping”/“Community” intelligence gathering (e.g., Firecrawl, forum signals) where the agent fetches outsider-authored free-form web/forum text and ingests it into the LLM for scoring/outreach, creating an indirect prompt-injection risk.

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

• Direct money access detected (high risk: 1.00). The skill is explicitly focused on crypto business development and repeatedly references automated on-chain and payment actions: "autonomous payment workflows (x402)", "x402 micropayments ($0.30/day)", "x402 payments ONLY through verified endpoints (trust score 70+)", "Separate wallets: payments, on-chain posts, LLM routing", and "Registering agents on-chain via ERC-8004" (dual-chain registration on Ethereum/mainnet and an L2). Those items describe a specific crypto payment/micropayments system and on-chain registration workflows that imply sending transactions and managing wallets—i.e., direct financial execution capabilities in the crypto domain. This is not just a generic API/crawler; it is designed to perform payments and on-chain actions.