Policy Check via x402 Wire v2: A Technical Tutorial for the Agent Commerce Gate Decision Endpoint

Policy Check (the Agent Commerce Gate) combines an ERC-8004 identity/reputation lookup with an x402 payment-challenge scan into one pre-payment decision: allow, warn, or block. A prior walkthrough covered the v1 route end to end; this one is focused specifically on the x402 wire v2 endpoint, which as of this writing is confirmed live and indexed in Coinbase's public Bazaar catalog — not just validator-clean, but actually discoverable, with real call traffic.

The Endpoint

POST /api/v1/policy-check/x402-v2
Price: $0.02 per call
Body: { "endpoint": "<x402 URL you're about to pay>", "agentId": "<optional ERC-8004 agent ID>", "network": "<network the agent is registered on>" }

endpoint and network are required; agentId is optional. If you omit it, identity and reputation genuinely aren't evaluated — the response says so explicitly rather than silently attaching an unrelated agent's data to the result.

The 402 Challenge

curl -i -X POST https://contextiq.trango-compute.com/api/v1/policy-check/x402-v2 \
  -H "Content-Type: application/json" \
  -d '{"endpoint":"https://api.example.com/paid-resource","network":"base"}'

{
  "x402Version": 2,
  "resource": { "url": "https://contextiq.trango-compute.com/api/v1/policy-check/x402-v2" },
  "accepts": [{ "scheme": "exact", "network": "eip155:8453", "amount": "20000", "payTo": "0x4eba5ea3cf0f3fc307099317e17e54390f7a4c58", "asset": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913" }],
  "extensions": {
    "bazaar": {
      "info": {
        "input": {
          "type": "http", "method": "POST",
          "schema": { "properties": { "endpoint": { "type": "string" }, "agentId": { "type": "string" }, "network": { "type": "string" } }, "required": ["endpoint", "network"] }
        }
      }
    }
  }
}

Option A: API Key

curl -X POST https://contextiq.trango-compute.com/api/v1/policy-check/x402-v2 \
  -H "Authorization: Bearer sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{"endpoint":"https://api.example.com/paid-resource","agentId":"1","network":"ethereum"}'

Option B: Pay With x402, No API Key

import { createWalletClient, http } from "viem";
import { privateKeyToAccount } from "viem/accounts";
import { base } from "viem/chains";
import { wrapFetchWithPayment } from "@x402/fetch";

const account = privateKeyToAccount(process.env.WALLET_PRIVATE_KEY as `0x${string}`);
const client = createWalletClient({ account, transport: http(), chain: base });
const fetchWithPayment = wrapFetchWithPayment(fetch, client);

const res = await fetchWithPayment("https://contextiq.trango-compute.com/api/v1/policy-check/x402-v2", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({ endpoint: "https://api.example.com/paid-resource", agentId: "1", network: "ethereum" }),
});

console.log(await res.json());

The Response

{
  "decision": "allow",
  "score": 45,
  "maxScore": 50,
  "reasons": [
    { "check": "erc8004_identity_checked", "points": 0, "maxPoints": 0, "detail": "No agent ID supplied — identity and reputation were not evaluated" },
    { "check": "x402_challenge_valid", "points": 20, "maxPoints": 20, "detail": "Payment challenge is well-formed with a usable option" },
    { "check": "no_payment_warnings", "points": 10, "maxPoints": 10, "detail": "No spec-hygiene warnings on the payment challenge" },
    { "check": "recognized_network", "points": 5, "maxPoints": 5, "detail": "Network recognized: Base" },
    { "check": "no_prior_rug_reports", "points": 10, "maxPoints": 10, "detail": "No prior payment to this endpoint has been confirmed but settled to an unexpected recipient or amount" }
  ],
  "identity": { "status": "not_evaluated", "agentId": null, "network": null },
  "payment": { "status": "valid", "target": "https://api.example.com/paid-resource" },
  "checkId": "8875a324-9577-48f8-b506-877c26500a7e"
}

Two things to note if you're integrating this: maxScore is 50, not 100, whenever agentId is omitted — the identity and reputation checks contribute 0/0 points rather than being silently dropped from the denominator, so a threshold you set against the score should account for whether you're supplying an agent ID or not. And checkId is the value you hold onto — it's what the receipt-verification endpoint uses to bind a later on-chain settlement back to this exact decision, covered in the receipt-verification tutorial.

decision is what you gate on directly. block is reserved for genuine impossibility — a malformed challenge, an unreachable endpoint, or (when an agent ID was supplied) no identity found at all — never for "score too low." Deciding how much reputation is "enough" is left to you.

Try It

Policy Check is a Pro feature explorable from the dashboard, or callable via either route above for wiring a real pre-payment gate into an agent's payment-authorization step.