Overview

Infinicept's Developer Documentation Model Context Protocol (MCP) server enables your agentic AI/LLM tools like Claude, Claude Code, ChatGPT, Gemini, and others, to interact directly with Infinicept's API Reference. It can look up API specs, pull endpoint details, and help generate code to streamline your integration with natural-language prompts.

This is just the first of Infinicept's MCP offerings. Stay tuned for more information coming soon.

MCP Server Tools

Connecting to the MCP Server

The Developer Documentation MCP server is available at https://developer.infinicept.com/mcp. Select a tab below for instructions to connect your preferred AI agent to this server:

  {
    "mcpServers": {
      "infinicept": {
        "url": "https://developer.infinicept.com/mcp"
      }
    }
  }

    claude mcp add --transport http -s user infinicept https://developer.infinicept.com/mcp

    claude mcp list

    {
      "mcpServers": {
        "infinicept": {
          "url": "https://developer.infinicept.com/mcp"
        }
      }
    }

    {
      "mcpServers": {
        "infinicept": {
          "url": "https://developer.infinicept.com/mcp"
        }
      }
    }

    {
      "mcpServers": {
        "infinicept": {
          "url": "https://developer.infinicept.com/mcp"
        }
      }
    }

Try It Now: Your First Query in Under 2 Minutes

Once you've connected to the MCP server using the instructions above, try this quick workflow to confirm everything is working and see the tools in action.

→ Step 1: Open your AI assistant (Claude, Claude Code, ChatGPT, Gemini, etc.) and start a new chat.

→ Step 2: Ask your AI what APIs are available.

Use Infinicept's documentation to list all available API specs.

You should see a response listing Infinicept's API specifications for New Accounts, Backoffice, and Orchestrate. This confirms the list-specs tool is connected and working.

→ Step 3: Explore an endpoint.

Use Infinicept's documentation to show me the full details on the GET endpoint for reviewing paginated transactions.

The response should return details including the endpoint path, parameters, authentication requirements, and a description all pulled directly from the live API specification.

→ Step 4: Generate a code snippet.

Use Infinicept's documentation to generate a Python code snippet for reviewing paginated transactions via the Backoffice API.

You should receive a working code snippet you can copy directly into your project.

More Example Outputs

Wondering what to expect? Here's what responses from the MCP server look like in practice:

Exploring webhooks

Prompt:

Use Infinicept's documentation to show me what webhook events are available with Backoffice.

What AI returns:

The Infinicept Backoffice API supports webhook subscriptions for key funding lifecycle events. Use the type field in your Subscribe to a Webhook request to specify which event you want to receive notifications for.

Webhook Event Types

Type Value	Event Name	Description
5	Submerchant Deposit Rejected	Triggered when a submerchant deposit has been rejected.
6	Funding Deposit Funded	Triggered when a funding deposit has been successfully funded.
7	Funding Deposit Rejected	Triggered when a funding deposit has been rejected.
8	Funding Deposit Partially Rejected	Triggered when a funding deposit has been partially rejected.

Example

To subscribe to the Funding Deposit Rejected webhook, pass 7 as the type value in the request body:

JSON

{
  "type": 7,
  "destinationURL": "https://your-domain.com/webhooks/funding-deposit-rejected",
  "authenticationMethod": {
    "method": 0,
    "username": "your-username",
    "password": "your-password"
  }
}

Getting endpoint details

Prompt:

Use Infinicept's documentation to give me the full details on the GET endpoint for reviewing paginated transactions.

What AI returns:

GET /{tenantId}/transactions/paginated

Authentication — requires x-AuthenticationKeyId and x-AuthenticationKeyValue headers.

Key query parameters include:

Parameter	Type	Description
SubmerchantId	integer	Filter by one or more submerchant IDs
TransactionType	string	Filter by type: Deposit, Refund, Chargeback, etc.
FundingStatus	string	Filter by status: New, Submitted, Funded, Held, etc.
DateFilter	string	All, Today, Yesterday, ThisWeek, DateRange, etc.
FromDate / ToDate	date-time	Required when DateFilter is set to DateRange
CurrentPage	integer	Page number for pagination
PageSize	integer	Results per page (1–1000, default 1000)

The response returns a transactions array with fields like id, transactionAmount, netAmount, fundingStatus, settlementDate, submerchantName, and more — along with pagination metadata (totalRequestedTransactions, currentPage, totalPages).

Generating a code snippet

Prompt:

Use Infinicept's documentation to generate a Python code snippet for retrieving paginated transactions via the Backoffice API.

What AI returns:

Here's a Python snippet for retrieving paginated transactions you can paste into your own environment.

Python

import requests

url = "https://api.infinicept.com/api/{tenantId}/transactions/paginated"

headers = {
    "accept": "application/json",
    "x-AuthenticationKeyId": "YOUR_KEY_ID",
    "x-AuthenticationKeyValue": "YOUR_KEY_VALUE"
}

params = {
    "DateFilter": "ThisMonth",
    "PageSize": 100,
    "CurrentPage": 1
}

response = requests.get(url, headers=headers, params=params)
print(response.json())

Support

💬 Have a question? Contact us on the Customer Support Portal.

🔹 Learn about updates in our Changelog.