code402.dev / for agents

Validation your agent can buy on its own

Every established bank-detail verification service requires a signup, a contract, and a procurement cycle. That is fine for a treasury team. It is impossible for an autonomous agent that needs to check a counterparty at 2am with no human in the loop. Your agent cannot sign a SaaS agreement — so it cannot buy from them at all.

That is the gap this exists for. No account, no API key to provision, no onboarding form. Your agent calls the endpoint, receives an HTTP 402 challenge, signs a payment authorization, and gets its answer. Per call, from a wallet it already controls.

What makes it agent-native

What these checks are — and are not. They are structural: checksum and format, deterministic and offline-computable. They do not prove an account exists, is open, or belongs to the name on an invoice. Our sanctions screen is address- and jurisdiction-level only over the OFAC SDN digital-currency address list — a no-match is not a sanctions clearance. Every response states its own scope. If your agent needs ownership or name verification, this is not that, and buying it expecting that would be a mistake.

Wire it into your agent

# MCP: add the remote server, then call any tool
https://hcrb.in/mcp

# REST: the same tools over HTTP, 402 in, receipt out
curl -X POST https://hcrb.in/v1/tools/iban-check/call \
  -H 'content-type: application/json' \
  -d '{"input":{"iban":"GB29NWBK60161331926819"}}'
Try a free callPay per call in USDC