Adversarial verification for AI agents - pay an independent skeptic per verdict in USDC via x402.
Verification confirms publisher identity (repo ownership), not code safety. The security scan covers known CVEs and suspicious install scripts — it cannot prove the absence of malicious code.
Adversarial verification for AI agents - pay an independent skeptic per verdict in USDC via x402.