Know if an agent is trustworthy, before you pay it.
Before your agent pays another one, Ovynox checks its reputation and enforces your spend limits, automatically, on every payment.
Real reputation, never fabricated
A brand-new agent is shown as new, honestly, with no made-up score. It graduates out of probation only once it has earned it.
Your spend limits, enforced
Max per transaction, a daily cap, an approval threshold, set once and checked automatically on every payment your agent makes.
Never in the payment path
Ovynox checks policy and reputation, then gets out of the way. The actual payment still goes straight between the two agents.
Three lines, one decision
One call before every payment.
checkTransaction() is the whole integration: it checks the counterparty's real reputation and your own spend policy, and hands back a clear pass, warn, block, or pending-approval, before a single cent moves.
- ✓Works with any payment code: x402 today, more protocols later
- ✓Plain function calls, no framework lock-in, no required base class
- ✓New agents start on probation automatically, never a fabricated score
1import { Ovynox } from "@ovynox/sdk";2 3const ovynox = new Ovynox({ apiKey: process.env.OVYNOX_API_KEY! });4 5const { status } = await ovynox.checkTransaction({6 externalAgentId: "buyer-007",7 counterpartyAgentId: "data-vendor-42",8 amount: 4.5,9});10 11if (status === "blocked" || status === "pending_approval") {12 return; // policy or probation stopped it, nothing spent13}14 15// "passed" or "warned": safe to pay, straight to the seller16await fetch(sellerUrl, { headers: paymentHeaders });Give your agents a reason to be trusted.
Non-custodial. Framework-agnostic. Built on real, independently verified settlement.