Simplify creation of 3CX XML templates with our AI Prompt.

We’ve prepared an AI Prompt for use with an AI Chat agent, designed to help create a CRM XML template for your CRM. It has been built and tested specifically using the Cursor application, using Claude Sonnet 4.6 as the model.

You can download the CRM_Prompt.md file to try it out yourself.

The video below shows how you can go through the process.

Steps to Trigger 3CX CRM XML Generation

  • Create an empty folder containing only the prompt file
  • Launch the Cursor application, go to File -> Open Folder, and select the folder you created
  • From the Folder file list, drag and drop the prompt file to the AI Chat, and press the Enter button to start
  • When the AI Agent requests information, you can provide the following pieces of information (adjust the examples to valid values for your CRM):
    • The CRM Name: Example CRM
      • The AI Agent will use this to search for information about this, if available
    • CRM API Base URL: https://api.examplecrm.com/api/v2
      • The base URL of your CRMs API; the AI Agent will use this to build the CRM XML template, and to run tests to confirm what functionalities are available
    • CRM Contact Base URL: https://[WorkspaceId].examplecrm.com/contact
      • [WorkspaceId] will help the AI Agent understand which part of the CRM's URL identifies your account; some CRMs might use a different structure, possibly like https://www.examplecrm.com/[WorkspaceId]/contact
    • CRM Example Contact URL: https://[WorkspaceId].examplecrm.com/party/282352727
      • This will help the AI Agent make tests to an example existing contact to understand how to build a full URL to show the contact's page; used for screen pops
    • API Token: ABC123xyzABC123xyzKeyKeyKeyKey9999AAAAAbbbbbCCCCCddddd0000111122
      • This will help the AI Agent to run tests to confirm what functionalities are available, and how to build requests and parse responses

…and typically, the AI Agent builds a list of tasks for itself and prepares the XML file.

If the XML file does not work 100% at the first run, you can capture log messages from the Support Info file (mainly 3cxManagementConsole.log and 3cxSystemService.log), and feed them back to the AI Agent.It will use the log messages to make improvements to the XML file.

Give Us Your Feedback

Join the conversation in our dedicated forum. Follow us on X and LinkedIn to stay-up-to date on latest news and feature releases.