Braintree MCP Server

What QuentinCody/braintree-mcp-server MCP server does

QuentinCody/braintree-mcp-server MCP server connects MCP-compatible AI clients with PayPal Braintree payment processing. Its interface is intentionally small: one tool checks connectivity and another sends GraphQL operations to Braintree. That GraphQL access can cover supported Braintree operations such as retrieving transaction details, reading customer records and payment method details, and creating a transaction by charging a payment method.

The project is unofficial and communicates with Braintree through its GraphQL API. It does not present a separate fixed catalog of customer-management actions in the documented interface; instead, the general GraphQL tool is used for the operations the Braintree API and account permissions support.

How it works

The server offers two transport modes. The STDIO implementation starts a server process for each client session and communicates over standard input and output. This mode is intended for Claude Desktop and other MCP clients that support STDIO. The SSE implementation runs as a persistent local web server and can accept multiple client connections. Its documented default endpoint is http://127.0.0.1:8001/sse, with the process listening on 127.0.0.1 port 8001 unless configured differently.

The braintree_ping tool provides a basic credential and connectivity check, returning pong when successful. The braintree_execute_graphql tool accepts a GraphQL query and variables, then returns the JSON response from Braintree. This lets an agent request fields such as transaction status, amount, currency, creation time, customer identity, and masked card details, or submit a charge mutation with an amount, order ID, payment method ID, and settlement option.

Setup and configuration

QuentinCody/braintree-mcp-server MCP server requires Python 3.13 or newer. After obtaining the repository, install its dependencies with pip install -e .. The documented server commands are python braintree_server.py for STDIO and python braintree_sse_server.py for SSE.

Place a .env file in the project root with these settings:

• BRAINTREE_MERCHANT_ID
• BRAINTREE_PUBLIC_KEY
• BRAINTREE_PRIVATE_KEY
• BRAINTREE_ENVIRONMENT

Set the environment to sandbox or production as appropriate. The credentials come from the Braintree Control Panel. The SSE host and port are configurable, while the defaults are localhost and port 8001.

Tools and capabilities

• Test whether the configured Braintree credentials work with braintree_ping.
• Run arbitrary Braintree GraphQL queries and mutations with braintree_execute_graphql.
• Fetch customer fields and available payment method details through GraphQL.
• Retrieve transaction status, monetary value, currency, and creation time.
• Create a payment transaction by charging a payment method.
• Serve MCP clients over STDIO or local SSE transport.

Limitations and notes

The repository is described as unofficial. Successful operations depend on valid Braintree credentials, network access to Braintree endpoints, account permissions, and applicable rate limits. The README documents local SSE binding rather than a hosted remote service. The project uses the MIT License and adds an academic citation requirement for publications, presentations, or reports based on the software; commercial and other non-academic use follows the stated MIT terms.