SmartFlow

Guide to connecting Incomaker MCP with AI agents

What is Incomaker MCP? Incomaker MCP (Model Context Protocol) is a server that allows AI agents (such as Claude, Codex, Cursor, etc.) to access your account's data and analytics. Simply state the requested tasks…

What is Incomaker MCP?

Incomaker MCP (Model Context Protocol) is a server that allows AI agents (such as Claude, Codex, Cursor, etc.) to access your account’s data and analytics. Simply state the requested tasks to the agent in natural language.

⚠️ Important note: In the current version, the connection works in read-only mode. The tool is therefore primarily intended for data analysis, reporting, and retrieving information. Currently, it cannot be used to create new campaigns, upload contacts, or modify settings.

Examples of what you can ask (reading and data analysis):

• “Show me the campaign performance for the last month.”
• “Which conversion SmartFlows have the highest revenue?”
• “How many contacts unsubscribed in February?”
• “Which products do customers most often buy together?”

Where can I find my API key?

For a successful connection, you will need your unique API key. You can get it as follows:
1. Log in to your Incomaker account. 2. Go to the E-shop info section. 3. At the bottom left of the page, you will find the API key item. 4. This token is exclusively yours and is secret. Copy it and replace the text YOUR_API_KEY with it in all the examples below.
The universal server address for all platforms is: https://mcp.incomaker.com/mcp

Setup for your platform

If you do not want to edit files manually, you can simply tell your AI tool that you want to install this MCP server and provide the given code with your API key. After successful installation, restart the application.

Opencode

For Opencode, the designation “remote” is used in the configuration instead of the standard “http” type. Add the following code snippet to your configuration file and do not forget to insert your API key:
{  “mcpServers”: {    “incomaker”: {      “type”: “remote”,      “url”: “https://mcp.incomaker.com/mcp”,      “headers”: {        “Authorization”: “Bearer YOUR_API_KEY”      }    }  }}

Claude Desktop

Open the configuration file for the Claude Desktop application:
• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json• Windows: %APPDATA%Claudeclaude_desktop_config.jsonAdd a new mcpServers section to it:
{  “mcpServers”: {    “incomaker”: {      “type”: “http”,      “url”: “https://mcp.incomaker.com/mcp”,      “headers”: {        “Authorization”: “Bearer YOUR_API_KEY”      }    }  }}
After saving the file, restart the Claude Desktop application. You will then find the available Incomaker tools under the MCP icon.

Claude Code (CLI)

In the root folder of your project (or in your home folder for global settings), create or edit the .mcp.json file. Paste the same configuration into it as shown above for Claude Desktop, and save the file.
The server will be detected automatically, which you can verify with the > health_check command.
Cursor
In the Cursor editor settings, go to Settings → MCP and add a new server with the following details:• Name: incomaker• Type: HTTP• URL: https://mcp.incomaker.com/mcp• Authorization Header: Bearer YOUR_API_KEY

OpenAI Codex (CLI)

You can add a new server by simply running this command in your terminal:

codex mcp add incomaker   –transport streamable_http   –url https://mcp.incomaker.com/mcp   –header “Authorization: Bearer YOUR_API_KEY”
You can then verify the configuration by entering the command codex mcp get incomaker.
ChatGPT
How to connect the Incomaker MCP server to standard ChatGPT:

Incomaker will soon be available in the ChatGPT app catalog, but until then, you can fully utilize the MCP in developer mode. From a user perspective, there is no difference. During the connection process, ChatGPT will warn you about security risks. As long as you follow this procedure, you face no security risks; however, it is slightly more labor-intensive than connecting a registered app. The registered Incomaker application is expected to appear in ChatGPT in the near future.

In ChatGPT, open Settings.

Open the Advanced settings menu (arrow on the right):

Turn on Developer Mode and return to the previous menu.

Enter the required details (see image). Name, Connection, and Authentication are absolutely critical. To maintain clarity and organization, we also recommend adding a Description and an Icon.
Use the following icon:

Then proceed to Advanced OAuth Settings.

In the OAuth Client ID field, enter the E-shop UUID, which you can find in Incomaker under E-shop Info in the bottom left corner:

You will also need to confirm the risk disclaimer. However, if you have followed this procedure exactly and have not disclosed your API keys or other codes to anyone, you have nothing to worry about.

You should now see the login screen for the Incomaker application. Click Log in via Incomaker:

Finally, a dialog box prompting you to enter your API key should appear. You can find this in Incomaker as well, on the same page where you just found the E-shop UUID.
After that, the MCP server should fully connect and respond to your queries. For example, you can try: “Give me the results of a few campaigns from the past month.”

Perplexity

Incomaker offers an MCP server that you can use to control it through various popular AI agents. Connecting them to Incomaker is described in this article. Here we’ll look at how to connect the Incomaker MCP server to Perplexity.

Incomaker will soon be available in Perplexity’s app catalog, but until then, the MCP can be fully used in developer mode. From a user’s perspective, there is no difference. During the connection process, Perplexity will warn you about security risks. If you follow this procedure correctly, you don’t face any security risks, but it is somewhat more laborious than connecting a registered app. The registered Incomaker app should appear in Perplexity in the near future.

In Perplexity, open Connectors.

Click Add Custom Connector:

Enter the required information into the following table.

Use the following icon:

Then proceed to Advanced.

In the Client ID field, enter the E-shop UUID, which you can find in Incomaker under E-shop Info in the bottom left:

You’ll also need to confirm the risk disclosure statement. However, if you’ve followed this procedure exactly and haven’t disclosed your API keys or other codes to anyone, you don’t need to worry.

You should now see the login screen for the Incomaker app. Click Sign in with Incomaker:

Expand it and click Add Connector in the top right corner.

Finally, a dialog box should appear prompting you to enter your API key. You’ll also find this in Incomaker, on the same page as the E-shop UUID from before.

After that, the MCP server should fully connect and respond to your queries. You could try, for example: “Give me the results of a few campaigns from the last month”

Was this article helpful?