Payman MCP Server

The Payman API MCP Server is a bridge between AI assistants and Payman AI’s payment infrastructure. By exposing Payman’s REST endpoints as MCP tools, the server lets conversational agents—such as Claude or Cursor—to perform real‑world financial operations entirely through natural language. This eliminates the need for developers to write custom SDK wrappers or manage API keys manually, enabling rapid prototyping of payment‑enabled workflows.

At its core, the server offers a suite of tools that cover every step in the payment lifecycle. Developers can first set‑api‑key to authenticate with Payman, then use create‑payee to register new recipients in one of three supported formats (TEST_RAILS, US_ACH, or CRYPTO_ADDRESS). Once a payee exists, the send‑payment tool sends funds with optional memos and amounts, while search‑payee lets agents locate recipients by name, contact details, or account number. Finally, check‑balance retrieves the current balance of the authenticated Payman account, allowing agents to confirm sufficient funds before initiating a transaction. Each operation is wrapped with comprehensive error handling so that conversational agents can gracefully report failures or prompt the user for corrective action.

The server’s value lies in its seamless integration into existing MCP workflows. Because it follows the MCP standard, any client that understands the protocol—whether a desktop assistant, web UI, or browser extension—can consume these tools without additional code. The server supports both standard I/O and Server‑Sent Events transports, giving developers flexibility to deploy it locally for desktop assistants or expose it as a web service for browser‑based tools. This duality ensures that the same set of payment capabilities can be used across platforms with minimal configuration.

Real‑world scenarios for Payman MCP include automated invoice payments, on‑the‑fly expense reimbursements, or even integrating cryptocurrency payouts into a chatbot that manages user rewards. For example, an e‑commerce support bot could ask a customer for payment details, create a payee on the fly, and transfer funds—all while maintaining conversational context. In enterprise settings, HR chatbots could automate salary disbursements to employees’ bank accounts or crypto wallets, reducing manual intervention and audit overhead.

Unique advantages of the Payman MCP Server stem from its security model and type‑specific payee handling. API keys are stored only in the active session, mitigating exposure risks, and the server distinguishes between test rails, ACH transfers, and crypto addresses at the tool level—allowing developers to write context‑aware logic that chooses the appropriate payment channel. Combined with robust error handling and a lightweight deployment model, this server provides developers with a reliable, secure, and developer‑friendly pathway to embed payment functionality directly into AI assistants.