Enables AI agents to run pre-purchase safety checks on shops, checkout and payment links, paid APIs, and x402 endpoints before paying, returning a pay, caution, or do_not_pay verdict with reasons.
Enables AI agents to scan untrusted web pages, emails, PDFs, and tool output for prompt injection across 17 languages, returning findings and cleaned text before the agent reads the content.
Enables AI agents to screen ERC-20 tokens on Base and Ethereum for honeypot and rug-pull risks, including upgradeable proxies, owner powers, and trading restrictions, before swapping or approving. It also provides best swap route quotes with approval targets and unsigned transactions for the agent to sign and send.