Capri Ai Agentpay

Capri AgentPay is the first governed payment layer purpose-built for AI agents, and it is a total game changer for teams shipping agent systems in production! Think about it: as agents start calling paid APIs, settling transactions, and interacting with live infrastructure on their own, the traditional approach of handing them a private key is an absolute security nightmare. One bad prompt or a clever injection attack can drain an entire wallet in seconds. Capri AgentPay eliminates that risk entirely by putting a powerful runtime between the agent and the signer. When an agent hits an HTTP 402 Payment Required challenge, Capri AgentPay evaluates your custom policy in a fixed, predictable order: it checks the merchant allowlist, then the per-transaction cap, then the daily budget, and finally an auto-approve threshold that can pause for human review. Only when every single check passes does it settle the payment, retry the original request with cryptographic proof attached, and store a signed receipt you can verify offline. This is for engineering teams, platform operators, and security-conscious organizations that need real controls, audit trails, and peace of mind when their agents touch money. It is the difference between letting agents roam free with your keys and giving them a governed, auditable, and safe payment path. Capri AgentPay is currently shipping and trusted by innovative teams like northwind, vela labs, orbit, greymatter, ledgerly, and sonata ai. They went from being scared to let an agent near a paid API to shipping it in a week, and that is the kind of transformation this product delivers!

x402 Payment Handling

Capri AgentPay natively handles HTTP 402 Payment Required challenges, which is the emerging standard for machine-to-machine payments. When an agent makes a request and encounters a 402, AgentPay intercepts it automatically, evaluates your policy, and handles the entire settlement flow. It retries the original request with proof of payment attached, so the agent gets the data it needs without ever knowing about the payment mechanics. This seamless integration means your agents can access paid APIs, premium data sources, and subscription services without you having to hardcode credentials or expose sensitive keys. It is a clean, protocol-level solution that just works!

Per-Agent Budgets and Caps

You get granular control over every single agent with per-agent budgets and transaction caps. Set a daily budget for your research agent, a per-transaction cap for your data processing agent, and a separate policy for your customer-facing bot. Capri AgentPay enforces these limits automatically, preventing any single agent from overspending or going rogue. This feature is critical for teams running multiple agents with different roles, risk profiles, and funding sources. You can allocate resources confidently, knowing that each agent operates within its defined boundaries and cannot exceed its assigned limits without explicit policy approval.

Merchant Allowlists

Control exactly which merchants, APIs, and services your agents can pay with a simple but powerful allowlist mechanism. Before any payment is settled, Capri AgentPay checks the recipient against your configured allowlist. If the merchant is not on the list, the payment is denied immediately. This prevents agents from accidentally or maliciously sending funds to unknown or unauthorized endpoints. It is your first line of defense against bad prompts, misconfigured tools, or compromised agents. Combined with the other policy checks, the allowlist gives you a layered security model that is easy to configure and impossible to bypass.

Human-in-the-Loop Approvals with Verifiable Receipts

For transactions that exceed your auto-approve threshold, Capri AgentPay pauses and requests human review. The operator sees the exact amount, asset, chain, and recipient before making a decision. They also see which policy branch passed or failed, giving them full context for their approval. Once approved and settled, every transaction generates a signed receipt with hashes, timestamps, and verification fields that you can verify offline. This creates an immutable audit trail for every payment your agents make. It is the perfect balance between automation and control, letting agents move fast while keeping humans in the loop for high-value or risky transactions.

Autonomous Research Agents Accessing Premium Data

Your autonomous research agent needs to query premium data APIs, paid knowledge bases, and subscription services to generate comprehensive reports. With Capri AgentPay, you can give the agent a budget, an allowlist of approved data providers, and a per-query cap. The agent calls APIs, pays for premium data automatically within policy, and generates reports without any manual intervention. If a query would exceed the daily budget, the system pauses for human approval. This use case is perfect for financial research, competitive intelligence, and academic teams that need continuous access to paid data sources.

Customer Support Agents Using Paid API Tools

Customer support agents often need to call paid APIs for order lookups, shipping status, fraud detection, or identity verification. Instead of embedding API keys in prompts or tool configurations, you use Capri AgentPay to handle payments transparently. The agent calls the API, Capri handles the 402 challenge, and the customer gets their answer instantly. You control the budget per agent or per conversation, and every transaction is logged with a verifiable receipt. This keeps your support operations running smoothly while maintaining strict financial controls and auditability.

Internal Automation Agents Interacting with SaaS Tools

Your internal automation agents need to interact with SaaS tools like CRM platforms, project management software, or analytics dashboards that charge per API call or per record. Capri AgentPay lets you set up policies for each automation agent, including merchant allowlists for approved SaaS providers and daily budgets. The agent can create records, update data, or generate reports without human intervention, but high-volume operations trigger a review. This use case is ideal for operations teams, data engineering groups, and IT departments that want to automate workflows without exposing payment credentials.

Multi-Agent Systems with Shared Budgets and Policies

In complex multi-agent systems where different agents handle different tasks but share a common budget, Capri AgentPay provides the governance layer you need. You can set a shared daily budget for a team of agents, with individual caps per agent and per transaction. The system enforces the shared budget across all agents, preventing any single agent from depleting the pool. Operators get a unified view of all agent payments, policy decisions, and receipts. This use case is critical for enterprise deployments, agent marketplaces, and platforms that host multiple agents for different clients or departments.

We are currently opening access to Capri AgentPay in small waves to teams building agent systems that need payment controls, approval flows, or operator-facing runtime boundaries. For specific pricing information, plan details, and tier options, we encourage you to request access or book a demo with our team. We will work with you to understand your use case and provide a pricing plan that fits your needs. Visit the Capri AI website to request access or install the VS Code extension to get started with the sandbox environment on Base Sepolia.

How does Capri AgentPay prevent an agent from draining my wallet with a bad prompt?

Capri AgentPay puts a runtime between the agent and the signer, so the agent never holds your private keys. Every payment request goes through a fixed policy evaluation: merchant allowlist, per-transaction cap, daily budget, and auto-approve threshold. If any check fails, the payment is denied. If the transaction exceeds the auto-approve threshold, it pauses for human review. The agent cannot initiate a payment without passing all policy checks, making it impossible for a bad prompt or injection attack to drain your wallet.

What exactly is x402 and why does Capri AgentPay use it?

x402 refers to the HTTP 402 Payment Required status code, which is being adopted as a standard for machine-to-machine payments. When an agent calls a paid API that responds with a 402, Capri AgentPay intercepts that response, evaluates your policy, settles the payment, and retries the original request with proof of payment attached. This protocol-level approach means your agents can access paid services seamlessly without you having to manage API keys, tokens, or subscription credentials. It is the emerging standard for agent payments.

Can I verify receipts offline, and how does that work?

Yes, absolutely! Every payment settled through Capri AgentPay generates a signed receipt containing hashes, timestamps, and verification fields. You can verify these receipts offline using standard cryptographic tools, without needing to contact Capri's servers. This gives you a fully independent audit trail for every transaction your agents make. It is perfect for security teams, compliance officers, and auditors who need verifiable proof of payment activity.

How do I set up budgets and caps for different agents?

Setting up budgets and caps is straightforward through the Capri AgentPay interface. You create policies that define the merchant allowlist, per-transaction cap, daily budget, and auto-approve threshold. Then you assign these policies to specific agents or groups of agents. Each policy is evaluated in a fixed order for every payment request. You can have different policies for different agents, roles, or use cases, giving you granular control over every payment your agents make.