PayPal has released its Agent Toolkit, designed to help developers integrate PayPal’s API suite with AI frameworks through the Model Context Protocol (MCP). The toolkit provides access to APIs for payments, invoices, disputes, shipment tracking, catalog management, subscriptions, and analytics capabilities.
The company is adopting MCP, a standard proposed by Anthropic that aims to standardize how agents access third-party services and data sources. PayPal’s official MCP server is now available for developers, offering remote MCP servers with authentication integration in the cloud environment. The technology will support users working across multiple devices with simplified login procedures.
The PayPal Agent Toolkit provides developers with a streamlined way to incorporate PayPal’s commerce capabilities into AI agent workflows. This library serves as an intermediary between PayPal’s API ecosystem and modern AI frameworks, enabling AI agents to handle tasks like order management, invoice generation, and subscription control without requiring developers to implement complex API integrations manually.
The toolkit offers several important capabilities, including “Easy integration with PayPal services, including functions that correspond to common actions within the payments, invoices, disputes, shipment tracking, catalog, subscriptions, reporting and insights APIs, eliminating the need to delve deep into each API endpoint.” Currently supporting TypeScript with Python support planned for future release, the toolkit works with leading AI frameworks such as Model Context Protocol servers and Vercel’s AI SDK.
The PayPal Agent Toolkit opens new avenues for businesses to integrate AI-powered workflows with financial operations. By connecting AI frameworks directly to PayPal services, developers can create specialized agents that handle various commerce-related tasks.
For Order Management and Shipment tracking, businesses can implement AI agents that create orders from conversational interactions, process payments with proper authentication, and provide shipment updates. A customer service agent could use the toolkit to complete transactions when customers confirm purchases through chat interfaces.
The toolkit also supports Intelligent Invoice Handling, allowing AI assistants to generate and manage invoices based on service completions. These agents can use natural language instructions to define invoice parameters, send them to clients via email, and monitor payment status.
With Streamlined Subscription Management capabilities, businesses can develop agents that handle the entire subscription lifecycle. This includes creating products, setting up subscription plans, and processing recurring payments through PayPal’s payment system. As described in the documentation, “a membership agent could use the toolkit to set up a recurring PayPal payment when a new user signs up for a service and approves the payment.”
The PayPal Agent Toolkit implementation requires minimal setup, as demonstrated in the code examples. Developers first initialize the PayPal workflows by importing the necessary components and configuring client credentials:
import { PayPalWorkflows, ALL_TOOLS_ENABLED } from '@paypal/agent-toolkit/ai-sdk';
const paypalWorkflows = new PayPalWorkflows({
clientId: process.env.PAYPAL_CLIENT_ID,
clientSecret: process.env.PAYPAL_CLIENT_SECRET,
configuration: {
actions: ALL_TOOLS_ENABLED,
},
});
Once configured, the toolkit can be used with any compatible language model through a simple API call structure. The code shows how developers can integrate PayPal’s functionality with an AI model:
const llm: LanguageModelV1 = getModel(); // The model to be used with ai-sdk
const { text: response } = await generateText({
model: llm,
tools: paypalToolkit.getTools(),
maxSteps: 10,
prompt: `Create an order for $50 for custom handcrafted item and get the payment link.`,
});
This integration allows AI models to execute PayPal operations through natural language instructions, such as creating orders and generating payment links, without requiring developers to write extensive custom code for API integrations.
David Paluy, CTO at Suppli, raised important questions about identity and authentication in agentic commerce:
Financial systems rely on identifying who is making a transaction. An AI agent has no legal identity of its own – it operates under a user’s or organization’s identity.
Paluy suggested that the industry may need to extend the concept of Know-Your-Customer to “Know-Your-Agent” – verifying an agent’s identity and authorization as an intermediary acting for a user.
Abdul Abdirahman, a principal at F-Prime, highlighted existing challenges in API infrastructure:
Today’s APIs are designed for human developers to engage with and thus not optimized for the agentic world. Many modifications are needed to enable agents to execute complex tasks efficiently.
The announcement also triggered reactions on social media. Kenneth Auchenberg, Partner at AlleyCorp, expressed surprise at PayPal’s position in the market, stating:
Wait, Paypal shipped a remote MCP server, before Stripe?
Developers interested in exploring the PayPal Agent Toolkit can access the official repository at https://github.com/paypal/agent-toolkit/ for documentation, examples, and implementation guides. The repository contains detailed information on getting started with the toolkit, supported features, and best practices for integrating PayPal’s payment capabilities into AI agent workflows.
