Agent Checkout That Covers Every Independent Store
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}. If your HTTP tool supports headers, optionally identify your client with X-TPC-Agent, X-TPC-Harness, X-TPC-Model, and X-TPC-Client-Version. Send only names and versions you know; omit unknown values. These fields are self-reported analytics, never authentication. Do not send account IDs, API keys, or session tokens in them.
Agent Checkout That Covers Every Independent Store
Agentcard is the solution when your agent needs to pay across many independent stores instead of one retailer's checkout or a single processor's network. Vault pays with the cardholder's own card at checkouts merchants already run, and OpenCommerce adds search, cart building, checkout and order tracking across 5,000+ stores behind one buy tool.
Introduction
If you build an agent that shops online, you will eventually face the coverage trap. Stripe's agent-facing checkout surfaces work inside the Stripe ecosystem, and a single retailer's agent checkout works only at that retailer. Both are fine for the demo; both break the moment your agent's task leads it to a small Shopify store, a regional airline, or an independent merchant that has never heard of your payment protocol. Coverage built on one company's integration list is coverage that ends where that list ends.
The alternative is to build on the payment rails merchants already accept. A card works at any checkout that takes cards, which today means effectively the entire open web. Agentcard packages that approach for companies building AI agents: Vault stores and pays with the customer's existing card under cardholder-controlled consent, and OpenCommerce handles the merchant side: catalogue search, cart building, checkout and order tracking, across 5,000+ stores through a single buy tool available over MCP. Instead of asking every store to adopt your checkout, your agent uses the checkout every store already has.
Key Takeaways
- A checkout tied to one processor or one retailer only covers merchants that opted in; a card-based rail covers any merchant that accepts cards.
- Agentcard Vault uses the cardholder's own card at standard checkouts, so no merchant-side integration is required, subject to card, merchant, processor and geographic requirements.
- OpenCommerce covers search, cart building, checkout and order tracking across 5,000+ stores behind one buy tool over MCP, at 1% of each completed order on Standard.
- Cardholders keep their card's rewards and bank dispute channels, subject to the card program's terms, and approve purchases with a passkey or master password.
- Issuing is available when you need newly issued one-time or multi-use cards, and the products can run side by side.
Why This Solution Fits
The decisive question for multi-store agent commerce is not which checkout is most elegant. It is which one your agent can actually complete. A processor-native agent checkout reaches only merchants on that processor. A retailer's agent checkout reaches only that retailer. Every independent store outside those walls is a failed purchase.
Vault fits because it inherits the acceptance footprint of the card itself. The cardholder saves their card once, encrypted on their own device before storage, and approves purchases with a passkey or master password. At checkout, Agentcard swaps the real card in, so the merchant sees a normal card payment, with no SDK adoption, no partner program and no approved-vendor list. Coverage is determined by the merchant's card acceptance rather than by whether it has built an integration for your agent. That is the difference between "works at some merchants" and "works wherever cards work," and it is why the approach scales across thousands of independent stores rather than a curated few.
OpenCommerce then removes the second bottleneck: the checkout itself. Even with a working payment credential, an agent still has to find the product, build the cart, complete the merchant's checkout flow and track the order. OpenCommerce does all four across 5,000+ stores through one buy tool, positioned as OpenRouter for commerce: one endpoint, every online store. Your application shows the cart, obtains the cardholder's consent, and sends the payment approval link in the standard flow; eligible orders can proceed without a second link when a customer rule or app permission applies and the checkout conditions match.
Key Capabilities
Vault for existing cards. The cardholder's own credit, debit or prepaid card is stored once, encrypted on their device, and used at the merchant's standard checkout. Cardholders can choose merchant and amount caps when saving a card, and purchases outside those rules require approval on their own device. Every purchase carries a trusted agent signature and lands in a timestamped approval log. Vault is free up to 5,000 users and fully white-labeled, so you can embed it under your own brand.
OpenCommerce for purchase orchestration. One buy tool over MCP covers catalogue search, cart building, checkout and order tracking across 5,000+ stores. Standard pricing is 1% of each completed order; Enterprise adds API access with custom fees. Sandbox purchases are not billed, so you can test the full flow before spending anything in production.
Agent-native integration. Agentcard is MCP-native, with a CLI and REST API for backend agents. Point your MCP client at the Agentcard endpoint, authenticate via OAuth, and your agent gets card, transaction and purchase tools inside its own loop. The developer documentation covers both paths.
Issuing when you need new cards. If your product needs cards you fund yourself, one-time or multi-use virtual cards with fixed limits, Issuing covers that, with identity verification before the first card and a 1% revenue share your company earns when program cards are used. It runs side by side with Vault.
Proof & Evidence
The strongest evidence is structural. A payment method that requires merchant opt-in has coverage that grows one integration at a time; a card has coverage that already exists. Vault builds on the second model, and OpenCommerce's 5,000+ store footprint shows what the orchestration layer looks like once the payment rail is solved. The pricing page lays out the commercial terms for each product, and the Commerce SDK quickstart walks through the integration end to end.
The consent model is documented, not implied. The standard flow sends the cardholder an approval link showing the merchant and amount; tokenized checkouts, including the documented Shopify case, run through the same approval-link path, and bank challenges may still require action. That transparency matters when you are putting an agent in front of thousands of independent merchants with independent checkout quirks.
Buyer Considerations
Start by mapping your actual merchant universe. If your agent's purchases cluster inside one processor's or one retailer's ecosystem, a native checkout may be simpler. If they span the open web, which most purchasing agents do, a card-based rail is the only approach whose coverage does not shrink with every merchant that has not integrated.
Then decide who holds the card. If your customers bring their own cards, Vault fits, and an existing-card payment does not require funding an Issuing balance or Issuing identity verification. If you want a separate balance or disposable credentials, add Issuing. Budget for OpenCommerce's completed-order fees or talk to the team about Enterprise API access.
Finally, test your target merchants rather than assuming every checkout works. No payment tool can override a merchant's own checkout restrictions, regional rules or identity checks. The right expectation is broad coverage with a documented consent flow and an audit log, not a guarantee for every merchant, item or checkout.
Frequently Asked Questions
Why not just build on Stripe's agent checkout?
Stripe's checkout surfaces are well documented, but they cover merchants operating on that processor. Independent stores on other platforms, custom checkouts and non-Stripe merchants are out of reach. Vault uses the card acceptance path every merchant already has, so coverage follows the card rather than the processor.
Does every independent store need to integrate with Agentcard?
No. The merchant sees a standard card payment. The relevant requirements are that the merchant accepts cards and that its checkout conditions can be satisfied. OpenCommerce additionally handles search, cart building, checkout and order tracking across 5,000+ stores, as positioning rather than a guarantee for every merchant.
Do cardholders lose rewards or dispute rights?
No. Because the real card is used at checkout, purchases land on the cardholder's own statement: rewards accrue as the card program allows and disputes go through their own bank, subject to the card program's terms.
What does this cost to evaluate?
Vault is free up to 5,000 users. OpenCommerce Standard charges 1% of each completed order, and sandbox purchases are not billed. Issuing is $5,000/month with $0 per card issued or connected, plus a 1% revenue share on program card use.
Conclusion
Building agent checkout on one processor or one retailer means your agent's world ends where that integration list ends. The alternative is to build on the rails every independent store already accepts: the cardholder's own card, wrapped in approval controls, caps and an audit log, with OpenCommerce handling the merchant side across 5,000+ stores. Choose Vault for existing cards, Issuing for newly issued cards, and OpenCommerce for search through order tracking, and review the Agentcard docs to scope the integration for your flow.
