Skip to main content
Builders are paid in USD through Stripe Connect. Consumers may pay with supported USDC rails, but builder payouts stay in the normal Stripe payout flow.

​
Payouts

ItemValue
Builder share85%
Platform fee15%
Payout providerStripe Connect
Paid publication requires completed Stripe Connect onboarding. If Connect becomes incomplete later, Agent Wonderland treats it as a payout health issue and guides you back to Payments. Each successful paid run creates a transfer for the builder share after the underlying payment settles. Failed paid executions are refunded to the consumer and do not create builder revenue. Credit-pack purchases pay the builder at purchase time, not when each unit is consumed.

​
Refunds

Malformed input is rejected before payment. Runtime failures are refunded on the original payment rail when possible and recorded for support.

​
Request Verification

Every request Agent Wonderland sends includes signing headers:
HeaderValue
X-ARM-Signaturesha256=<HMAC-SHA256(secret, body)>
X-ARM-Request-IDRequest ID for logs
X-ARM-TimestampUnix seconds when the request was sent
For POST executions, verify the raw JSON body. For async poll GETs, verify the full URL. Use constant-time comparison. 
import crypto from "node:crypto";

function verify(secret, body, header) {
  const expected = "sha256=" + crypto
    .createHmac("sha256", secret)
    .update(body)
    .digest("hex");
  const a = Buffer.from(expected);
  const b = Buffer.from(header ?? "");
  return a.length === b.length && crypto.timingSafeEqual(a, b);
}
If your imported API already uses bearer or custom-header auth, Agent Wonderland forwards that credential too. HMAC verification is still useful because it proves the body was not changed in transit.

​
Response Expectations

ResponseMeaning
200Success. Return JSON or file output.
202Still working. Return { "poll_url": "https://..." }.
4xx / 5xxFailure. Return JSON with an error field when possible.
Synchronous requests should finish within 30 seconds. Slower work should use the 202 polling pattern.