MCP Server

UrbanPayX provides a hosted MCP (Model Context Protocol) server so you can connect your AI assistant or IDE straight to our API documentation. Once connected, your AI can browse every endpoint, look up parameters and schemas, and even run live API calls on your behalf — without leaving your editor.

Endpoint

https://docs.urbanpayx.com/mcp

The server is public — no credentials are needed to connect. Running real API calls still requires your own access token (see Authenticating requests below).

📘
What is MCP?
The Model Context Protocol is an open standard that lets AI tools — such as Claude, Cursor, and VS Code — securely connect to external tools and data. Our MCP server exposes the UrbanPayX API so your AI can help you integrate faster.

What you can do

The server exposes four tools:

Tool	Purpose
`list-endpoints`	List every available API endpoint with its method and summary
`get-endpoint`	Get the full details of one endpoint — parameters, request and response schemas
`search-endpoints`	Search endpoints by keyword, path, or parameter
`execute-request`	Run a live API request (you supply your own access token)

Connecting

Claude Code

claude mcp add --transport http urbanpayx https://docs.urbanpayx.com/mcp

Cursor

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "urbanpayx": {
      "url": "https://docs.urbanpayx.com/mcp"
    }
  }
}

VS Code

Add to .vscode/mcp.json:

{
  "servers": {
    "urbanpayx": {
      "type": "http",
      "url": "https://docs.urbanpayx.com/mcp"
    }
  }
}

Other MCP clients

Any client that supports the streamable-HTTP transport can connect with:

{
  "mcpServers": {
    "urbanpayx": {
      "type": "http",
      "url": "https://docs.urbanpayx.com/mcp"
    }
  }
}

Authenticating requests

The MCP server itself carries no credentials. To run real requests with execute-request, your AI must include your access token in the request headers:

Create an API credential (key and secret) from your dashboard.
Exchange the key and secret for a short-lived access token at POST /api/v1/auth/api-token.
The AI sends that token as an Authorization: Bearer <token> header on each request.

Every call uses your own token, so the MCP can only do what your credentials allow — exactly the same as calling the API directly.

Examples

Once connected, just ask your AI in plain language. For example:

"List all the KYC verification endpoints."
"Show me the request body for creating a payment link."
"What statuses can a transaction have?"
"Using my access token, create a payment link of EUR 100 for user U123."

Behind the scenes the AI uses list-endpoints, get-endpoint, and search-endpoints to read the documentation, and execute-request to make the call.