PayPal Agent Toolkit
Integrate PayPal APIs with popular agent frameworks using function calling.
Key Features
Use Cases
README
PayPal Agent Toolkit
The PayPal Agent Toolkit enables popular agent frameworks including OpenAI's Agent SDK, LangChain, Vercel's AI SDK, and Model Context Protocol (MCP) to integrate with PayPal APIs through function calling. It includes support for TypeScript and is built on top of PayPal APIs and the PayPal SDKs.
Available tools
The PayPal Agent toolkit provides the following tools:
Invoices
create_invoice: Create a new invoice in the PayPal systemlist_invoices: List invoices with optional pagination and filteringget_invoice: Retrieve details of a specific invoicesend_invoice: Send an invoice to recipientssend_invoice_reminder: Send a reminder for an existing invoicecancel_sent_invoice: Cancel a sent invoicegenerate_invoice_qr_code: Generate a QR code for an invoice
Payments
create_order: Create an order in PayPal system based on provided detailsget_order: Retrieve the details of an orderpay_order: Process payment for an authorized ordercreate_refund: Process a refund for a captured payment.get_refund: Get the details for a specific refund.
Dispute Management
list_disputes: Retrieve a summary of all open disputesget_dispute: Retrieve detailed information of a specific disputeaccept_dispute_claim: Accept a dispute claim
Shipment Tracking
create_shipment_tracking: Create a shipment tracking recordget_shipment_tracking: Retrieve shipment tracking informationupdate_shipment_tracking: Update shipment tracking information
Catalog Management
create_product: Create a new product in the PayPal cataloglist_products: List products with optional pagination and filteringshow_product_details: Retrieve details of a specific product
Subscription Management
create_subscription_plan: Create a new subscription planlist_subscription_plans: List subscription plansshow_subscription_plan_details: Retrieve details of a specific subscription plancreate_subscription: Create a new subscriptionshow_subscription_details: Retrieve details of a specific subscriptionupdate_subscription: update an existing subscriptioncancel_subscription: Cancel an active subscription
Reporting and Insights
list_transactions: List transactions with optional pagination and filteringget_merchant_insights: Retrieve business intelligence metrics and analytics for a merchant
TypeScript
Installation
You don't need this source code unless you want to modify the package. If you just want to use the package run:
npm install @paypal/agent-toolkit
Requirements
- Node 18+
Usage
The library needs to be configured with your account's client id and secret which is available in your PayPal Developer Dashboard.
The toolkit works with Vercel's AI SDK and can be passed as a list of tools. For more details, refer our examples
import { PayPalAgentToolkit } from '@paypal/agent-toolkit/ai-sdk';
const paypalToolkit = new PayPalAgentToolkit({
clientId: process.env.PAYPAL_CLIENT_ID,
clientSecret: process.env.PAYPAL_CLIENT_SECRET,
configuration: {
actions: {
invoices: {
create: true,
list: true,
send: true,
sendReminder: true,
cancel: true,
generateQRC: true,
},
products: { create: true, list: true, update: true },
subscriptionPlans: { create: true, list: true, show: true },
shipment: { create: true, show: true, cancel: true },
orders: { create: true, get: true },
disputes: { list: true, get: true },
},
},
});
To use sandbox mode, add context within your configuration.
configuration: {
context: {
sandbox: true,
}
}
Initializing the Workflows
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,
},
});
Usage
Using the toolkit
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.`,
});
PayPal Model Context Protocol
The PayPal Model Context Protocol server allows you to integrate with PayPal APIs through function calling. This protocol supports various tools to interact with different PayPal services.
Running MCP Inspector
To run the PayPal MCP server using npx, use the following command:
npx -y @paypal/mcp --tools=all PAYPAL_ACCESS_TOKEN="YOUR_ACCESS_TOKEN" PAYPAL_ENVIRONMENT="SANDBOX"
Replace YOUR_ACCESS_TOKEN with active access token generated using these steps: PayPal access token. Alternatively, you could set the PAYPAL_ACCESS_TOKEN in your environment variables.
Custom MCP Server
You can set up your own MCP server. For example:
import { PayPalAgentToolkit } from “@paypal/agent-toolkit/modelcontextprotocol";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
const orderSummary = await paypalWorkflows.generateOrder(
llm,
transactionInfo,
merchantInfo,
);
const server = new PayPalAgentToolkit({
accessToken: process.env.PAYPAL_ACCESS_TOKEN
});
async function main() {
const transport = new StdioServerTransport();
await server.connect(transport);
console.error("PayPal MCP Server running on stdio");
}
main().catch((error) => {
console.error("Fatal error in main():", error);
process.exit(1);
});
Usage with MCP host (Claude Desktop/Cline/Cursor/Github Co-Pilot)
This guide explains how to integrate the PayPal connector with Claude Desktop.
Prerequisites
- Claude Desktop application installed
- installing Node.js locally
Installation Steps
1. Install Node.js
Node.js is required for the PayPal connector to function:
- Visit the Node.js official website, download and install it.
- Requirements: Node 18+
2. Configure PayPal Connector with MCP host (Claude desktop / Cursor / Cline)
We will show the integration with Claude desktop. You can use your favorite MCP host.
- Open Claude Desktop
- Navigate to Settings
- Find the Developer or Advanced settings section
- Locate the external tools or connectors configuration area
- Add the following PayPal connector configuration to this ~/Claude/claude_desktop_config.json:
{
"mcpServers": {
"paypal": {
"command": "npx",
"args": [
"-y",
"@paypal/mcp",
"--tools=all"
],
"env": {
"PAYPAL_ACCESS_TOKEN": "YOUR_PAYPAL_ACCESS_TOKEN",
"PAYPAL_ENVIRONMENT": "SANDBOX"
}
}
}
}
Make sure to replace YOUR_PAYPAL_ACCESS_TOKEN with your actual PayPal Access Token. Alternatively, you could set the PAYPAL_ACCESS_TOKEN as an environment variable. You can also pass it as an argument using --access-token in "args"
Set PAYPAL_ENVIRONMENT value as either SANDBOX for stage testing and PRODUCTION for production environment.
- Save your configuration changes
3. Test the Integration
- Quit and restart Claude Desktop to apply changes
- Test the connection by asking Claude to perform a PayPal-related task
- Example: "List my PayPal invoices"
Environment Variables
The following environment variables can be used:
PAYPAL_ACCESS_TOKEN: Your PayPal Access TokenPAYPAL_ENVIRONMENT: Set toSANDBOXfor sandbox mode,PRODUCTIONfor production (defaults toSANDBOXmode)
This guide explains how to generate an access token for PayPal API integration, including how to find your client ID and client secret.
Prerequisites
- PayPal Developer account (for Sandbox)
- PayPal Business account (for production)
Finding Your Client ID and Client Secret
-
Create a PayPal Developer Account:
- Go to PayPal Developer Dashboard
- Sign up or log in with your PayPal credentials
-
Access Your Credentials:
- In the Developer Dashboard, click on Apps & Credentials in the menu
- Switch between Sandbox and Live modes depending on your needs
-
Create or View an App:
- To create a new app, click Create App
- Give your app a name and select a Business account to associate with it
- For existing apps, click on the app name to view details
-
Retrieve Credentials:
- Once your app is created or selected, you'll see a screen with your:
- Client ID: A public identifier for your app
- Client Secret: A private key (shown after clicking "Show")
- Save these credentials securely as they are required for generating access tokens
- Once your app is created or selected, you'll see a screen with your:
Generating an Access Token
Using cURL
curl -v https://api-m.sandbox.paypal.com/v1/oauth2/token \\
-H \"Accept: application/json\" \\
-H \"Accept-Language: en_US\" \\
-u \"CLIENT_ID:CLIENT_SECRET\" \\
-d \"grant_type=client_credentials\"
Replace CLIENT_ID and CLIENT_SECRET with your actual credentials. For production, use https://api-m.paypal.com instead of the sandbox URL.
Using Postman
- Create a new request to
https://api-m.sandbox.paypal.com/v1/oauth2/token - Set method to POST
- Under Authorization, select Basic Auth and enter your Client ID and Client Secret
- Under Body, select x-www-form-urlencoded and add a key
grant_typewith valueclient_credentials - Send the request
Response
A successful response will look like:
{
"scope": "...",
"access_token": "Your Access Token",
"token_type": "Bearer",
"app_id": "APP-80W284485P519543T",
"expires_in": 32400,
"nonce": "..."
}
Copy the access_token value for use in your Claude Desktop integration.
Token Details
- Sandbox Tokens: Valid for 3-8 hours
- Production Tokens: Valid for 8 hours
- It's recommended to implement token refresh logic before expiration
Using the Token with Claude Desktop
Once you have your access token, update the PAYPAL_ACCESS_TOKEN value in your Claude Desktop connector configuration:
{
"env": {
"PAYPAL_ACCESS_TOKEN": "YOUR_NEW_ACCESS_TOKEN",
"PAYPAL_ENVIRONMENT": "SANDBOX"
}
}
Best Practices
- Store client ID and client secret securely
- Implement token refresh logic to handle token expiration
- Use environment-specific tokens (sandbox for testing, production for real transactions)
- Avoid hardcoding tokens in application code
Disclaimer
AI-generated content may be inaccurate or incomplete. Users are responsible for independently verifying any information before relying on it. PayPal makes no guarantees regarding output accuracy and is not liable for any decisions, actions, or consequences resulting from its use.
Star History
Repository Owner
Organization
Repository Details
Programming Languages
Tags
Join Our Newsletter
Stay updated with the latest AI tools, news, and offers by subscribing to our weekly newsletter.
Related MCPs
Discover similar Model Context Protocol servers
Mifos MCP - Model Context Protocol (MCP)
Enabling AI agents to access and operate on financial data in the Mifos X ecosystem using Model Context Protocol.
Mifos MCP provides a Model Context Protocol (MCP) interface tailored for the Mifos X ecosystem, empowering AI agents to interact with financial data and perform operations within the platform. It offers a Java (Quarkus) implementation, with easy configurability via environment variables and native executable support. Developers can test and debug deployments with the MCP Inspector, and detailed instructions ensure smooth setup and integration. Comprehensive examples demonstrate practical banking and financial automation workflows.
- ⭐ 17
- MCP
- openMF/mcp-mifosx
Mastra
A TypeScript framework for building scalable AI-powered agents and applications.
Mastra is a modern TypeScript-based framework designed for developing AI-powered applications and autonomous agents. It offers model routing to integrate over 40 AI providers, a graph-based workflow engine, advanced context management, and production-ready tools for observability and evaluation. Mastra features built-in support for authoring Model Context Protocol (MCP) servers, enabling standardized exposure of agents, tools, and structured AI resources via the MCP interface.
- ⭐ 18,189
- MCP
- mastra-ai/mastra
MCP XRPL Server
Unified Model Context Protocol server for seamless AI-driven interaction with the XRP Ledger.
MCP XRPL Server provides a comprehensive Model Context Protocol implementation for the XRP Ledger, enabling AI agents to access blockchain services seamlessly across MainNet, TestNet, and DevNet. It offers an extensive suite of features including account management, token transfers, NFT operations, DIDs, AMM operations, and oracle interactions, all via a standardized MCP interface. The server supports context-driven interactions, making XRPL functionality readily accessible for AI-powered agents.
- ⭐ 5
- MCP
- RomThpt/mcp-xrpl
Agentset MCP
Open-source MCP server for Retrieval-Augmented Generation (RAG) document applications.
Agentset MCP provides a Model Context Protocol (MCP) server designed to power context-aware, document-based applications using Retrieval-Augmented Generation. It enables developers to rapidly integrate intelligent context retrieval into their workflows and supports integration with AI platforms such as Claude. The server is easily installable via major JavaScript package managers and supports environment configuration for namespaces, tenant IDs, and tool descriptions.
- ⭐ 22
- MCP
- agentset-ai/mcp-server
Braintree MCP Server
Enables AI assistants to securely access and manage Braintree payment processing via MCP.
Braintree MCP Server implements the Model Context Protocol (MCP) to provide AI assistants with structured, direct access to PayPal Braintree's payment services. It supports both STDIO and Server-Sent Events (SSE) transports, facilitating integration with AI clients and standalone web-based deployments. The server allows execution of GraphQL queries against Braintree, supports multi-client access, and handles authentication through environment-based configuration. Designed for secure, extensible, and automated payment operations in AI-driven workflows.
- ⭐ 3
- MCP
- QuentinCody/braintree-mcp-server
Remote-MCP
A type-safe, bidirectional system for remote Model Context Protocol communication.
Remote-MCP provides a simple and secure way to enable remote access to and centralized management of model contexts using the Model Context Protocol. It bridges local MCP clients with remote MCP servers, supporting a modular architecture via tRPC over HTTP. The tool allows integration both as a client and a server, catering to real-time remote access needs ahead of the official MCP roadmap.
- ⭐ 200
- MCP
- ssut/Remote-MCP
Didn't find tool you were looking for?