Guides to building agents that handle money

Learn how agent payments work, where controls belong, and how to test them safely.

The essentials

Start with the core concepts, then move into policies, approvals, and integration patterns.

Market overview

How AI agents buy, book, refund, pay APIs, and trigger payouts under policy.

Category guide

How to give AI agents spend limits, guardrails, and approval controls so they can pay for work without overspending.

Spend policies

Budgets, policies, approvals, wallets, counterparties, and audit trails for agents that spend money.

Micropayments

How to keep paid API calls, data access, and machine-speed service usage inside policy before an agent pays.

Product, finance, operations, and engineering teams each touch a different part of the path from agent action to approved spend.

Shape payment flows where agent actions, user trust, and approval moments are clear.

Keep spend limits, approval evidence, and reconciliation visible from request to settlement.

Build policy checks, budgets, wallets, and x402 handling into the request path.

Policies define the boundary, wallets hold authority, approvals add judgment, and audit trails make each payment reviewable.