PortOne

You set up and run the PortOne MCP Server to enable seamless querying and actions for PortOne tools. This server lets you retrieve channel lists, add test channels, fetch sub-store data, and view payment histories, all through your MCP client integrations. It’s designed to work with supported AI tools and helps you access PortOne features from your MCP ecosystem.

How to use

Connect your MCP client to PortOne MCP Server to access the available tools. Once registered, you can perform the following actions through your MCP client: - View the list of channels. - Add a test channel for quick validation. - Retrieve sub-stores and related data. - Look up payment histories. When using console-enabled features, you will need to sign in to the PortOne Console through your browser.

How to install

Prerequisites you must have before installing:
- Node.js 22.6.0 or newer
- An MCP client that supports loading an MCP server (or the environment where you manage MCP servers)
- Optional: pnpm is recommended for package management, but npm can also be used.

Additional sections

MCP server registration can be done by adding a local stdio configuration that points to the MCP server package. The following example shows how to register PortOne MCP Server via the MCP client settings. The command uses npx to fetch the latest PortOne MCP Server package and register it under the name portone_mcp.

"mcpServers": {
  "portone_mcp": {
    "command": "npx",
    "args": [
      "-y",
      "@portone/mcp-server@latest"
    ]
  }
}

Security and maintenance notes

When you use third-party AI services, be aware that API responses may be transmitted to the AI provider. PortOne MCP Server implements protections to identify and redact personal data in responses, but other data may be stored or processed in accordance with service policies.

Notes on installation via DXT and activation

You can install MCP Server using Desktop Extensions (DXT) by downloading the latest portone-mcp-server.dxt from the release channel and installing it with your preferred AI tool. After installation, restart the tool to confirm the MCP server is registered.

Migration and upgrades

If you previously used a different MCP setup (for example a Python-based server), migrate to the TypeScript version by updating the MCP configuration to load the new package and ensure Node.js is upgraded to 22.6.0 or newer. After updating the settings, restart your AI tool to apply changes.

Available tools

listChannels

Fetches and returns the list of available channels from the PortOne MCP Server so your client can display or process channel data.

addTestChannel

Creates a test channel through the PortOne MCP Server to validate end-to-end connectivity and data flow within your MCP client.

listSubStores

Queries and returns sub-store information, enabling you to explore nested store structures and related data within the PortOne ecosystem.

lookupPayments

Retrieves payment histories and related records for reconciliation and auditing within your MCP client.

consoleLogin

Facilitates console login actions for enabling browser-based PortOne Console interactions when using the MCP-enabled features.