Zen7-Payment-Agent

Zen7 Payment Agent acts as a bridge that allows AI assistants to handle real-world financial transactions autonomously. In simple terms, it gives AI agents their own digital wallets, enabling them to browse, order, and pay for goods or services on behalf of a user. Instead of just talking about a purchase, an AI equipped with this tool can actually complete the checkout process, making "agentic commerce" a reality where bots can interact with payment systems just as easily as they process text. The system is built on a sophisticated multi-agent architecture designed to handle the complexities of decentralized finance. It utilizes a central "host agent" to coordinate between specialized sub-agents that manage distinct parts of the transaction lifecycle: a payer agent for wallet verification, a settlement agent for executing on-chain transactions, and a payee agent for confirming receipt. By supporting both custodial and non-custodial models, it provides flexibility in how private keys are managed, while implementing gasless operations and passwordless authentication to ensure a smooth, high-frequency transaction experience. For developers integrating with LLMs, this MCP provides a robust protocol for intent recognition and automated settlement. It implements the Model Context Protocol (MCP) via the FastMCP framework, allowing models to invoke complex blockchain functions—like EIP-712 typed data signing and permit execution for tokens like USDC and DAI—through simple tool calls. By handling the intricate plumbing of blockchain network configurations and transaction monitoring, it allows developers to focus on building intelligent agent behaviors while ensuring that financial exchanges remain secure, verifiable, and decentralized.

Category: CRM, ERP & E-commerce

Tags: blockchain, defi, payments, smart-contracts, usdc, wallet

Visit Zen7-Payment-Agent

How to install and configure Zen7-Payment-Agent

  1. Install Python 3.13+, the uv package manager, and Git according to the repository setup documentation. 2. Clone the repository from https://github.com/Zen7-Labs/Zen7-Payment-Agent and configure your blockchain testnet environment and test wallet accounts. 3. Set up the local PostgreSQL database using SQLModel to store payment, settlement, intent, and audit event records. 4. Start the MCP server component located in the mcp_server directory, which runs using the FastMCP framework over Server-Sent Events (SSE) on port 8015 by default. 5. Register the SSE server URL (http://localhost:8015/sse) in your MCP client configuration, such as Claude Desktop or Cursor, following the repository README guides.

What you can do with Zen7-Payment-Agent

Use Case 1: Autonomous Procurement for AI Business Agents Problem: Modern businesses use AI to source goods and services, but the final payment step usually requires a human to log in, authorize a credit card, or manually sign a crypto transaction, which breaks the end-to-end automation. Solution: Zen7-Payment-Agent enables "permissionless authorization" and LLM-driven intent recognition. An AI buyer can autonomously negotiate and settle a transaction with an AI seller using secure, automated encrypted payments. Example: A "Project Manager Agent" identifies a need for 500 units of cloud storage. It contacts a "Storage Provider Agent," agrees on a price in USDC, and the Zen7 payer_agent generates an EIP-712 signature to authorize the transfer. The transaction is completed without the human manager needing to manually approve the gas or the signature.

Use Case 2: Pay-per-Query Marketplace for Specialized AI Skills Problem: Most AI services require expensive monthly subscriptions. If one AI agent needs a single specialized task from another (e.g., a "Legal Agent" needing a one-time translation from a "Translator Agent"), setting up a recurring payment or manual invoice is inefficient. Solution: The MCP's support for A2A (Agent-to-Agent) protocols and high-frequency transactions allows for instant settlement for individual tasks. It uses EIP-712 "permit" signatures to allow tokens like USDC or DAI to be transferred based on specific intent. Example: An AI Research Agent needs a summary of a massive dataset. It sends the request to a Summarization Agent. Upon delivery, the Zen7 settlement_agent automatically triggers a 0.50 USDC payment from the Researcher’s wallet to the Summarizer’s wallet, ensuring the specialized agent is paid instantly for that specific query.

Use Case 3: Gasless E-commerce for Non-Crypto Users Problem: Users often have stablecoins (like USDC) but no native blockchain tokens (like ETH) to pay for "gas" fees. This creates a "dead-end" where users cannot spend their money without first buying ETH from an exchange. Solution: Zen7 supports gasless operations through its `services

execute_permit.py` implementation. It allows the payment agent to handle the blockchain interaction costs, letting the user pay only the item price using their stablecoins. Example: A customer wants to buy a digital book using a mobile wallet. They have 20 USDC but 0 ETH. The Zen7 Payment Agent recognizes the intent, uses a permit signature to authorize the 20 USDC transfer, and executes the transaction on the Base network without requiring the user to own or understand "gas fees." ###…

Key facts

  • Open Source
  • https://github.com/Zen7-Labs/Zen7-Payment-Agent
  • CRM, ERP & E-commerce, Finance, Crypto & Payments
  • blockchain, defi, payments, smart-contracts, usdc, wallet

Part of MCP Servers

Related MCP servers

  • MCP Microsoft Office Bridge — MCP Microsoft Office Bridge is an MCP server that connects AI assistants like Claude to Microsoft 365 services through the…
  • MCP Mempool — MCP Mempool is an MCP server that exposes the mempool.space WebSocket and REST APIs to AI agents, LLM assistants, and…
  • MCP NIF.PT — MCP NIF.PT is an MCP server that connects LLM clients to the Portuguese NIF.PT public API to retrieve and analyze…
  • MCP Options Order Flow Server — MCP Options Order Flow Server is an MCP server that provides real-time options order flow data, institutional bias tracking, and…
  • MCP Open Library — MCP Open Library is an MCP server that connects AI assistants to the Open Library catalogue API to retrieve book,…
  • MCP Orlen Wholesale Price — MCP Orlen Wholesale Price is an MCP server that provides access to Polish Orlen fuel wholesale pricing data directly inside…

What is Zen7 Payment Agent?

Zen7 Payment Agent is an MCP server and payment infrastructure implementation designed for decentralized agentic commerce. It connects AI assistants to EVM and Solana networks to process cryptocurrency payments, verify balances, and manage order settlements. It uses a host coordinator alongside specialized sub-agents for payers, settlements, payees, orders, and token allowances.

What can Zen7 Payment Agent do?

The server enables AI models to perform automated crypto settlements, inspect multi-chain token authorizations, generate EIP-712 typed signatures or Ed25519 signatures, and monitor on-chain transaction execution. It also supports gasless transactions through permit executions for stablecoins like USDC and DAI across custodial and non-custodial wallet architectures.

Which blockchain networks are supported by Zen7 Payment Agent?

Zen7 Payment Agent supports EVM-compatible testnets including Ethereum Sepolia, Base Sepolia, Polygon Amoy, and BNB Chain Testnet, as well as Solana Devnet and Testnet. For tokens, it works with EVM USDC and DAI standards, alongside Solana SPL tokens, integrating with wallets like MetaMask, Coinbase Wallet, and Phantom Wallet.

Which MCP clients work with Zen7 Payment Agent?

Zen7 Payment Agent works with any Model Context Protocol client that supports Server-Sent Events SSE connections, including Claude Desktop, Cursor, and custom agent applications. The server exposes tools like proceed_payment_and_settlement_detail_info on port 8015, which client models invoke during conversational payment and purchasing workflows.

Is Zen7 Payment Agent open source?

Yes, Zen7 Payment Agent is open source software distributed under the Apache License Version 2.0. The complete source code, including sub-agent implementations, task managers, database schemas, and multi-chain service handlers, is accessible to developers for review, extension, and deployment on GitHub.

  • AI Tools
  • Categories
  • Industries
  • CLI Coding Agents
  • MCP Servers
  • MCP Categories