Using the MCP server

After it is configured, the MCP server enables natural language interaction with Decision Designer, allowing you to manage decision automations, models, and business rules through conversational commands.

What you can do

The MCP server supports comprehensive decision management operations:

Category	Available operations
Decision automations	List decision automations Create decision automations
Decision services	List decision services within a decision automation Create decision services within a decision automation
Sample decision services	Browse available samples Import samples into decision automations
Data models	List data models Create data models Update data models
Decision models	List decision models Create decision models Update decision models
Business rules	List business rules Create business rules Update business rules Delete business rules
Decision tables	List decision tables Create decision tables Update decision tables Delete decision tables
Test data sets	List test data sets Create test data sets Update test data sets
Decision operations	List decision operations Create decision operations
Dependencies	Manage dependencies between decision artifacts
Error report	Check data models and decision models for errors
Run	Run decision models by using test data

After you perform operations with your AI assistant, refresh Decision Designer to see your changes.

Remember: The accuracy and quality of results depend heavily on your assistant's capabilities. Use the most recent AI models for optimal performance.

Common use cases and examples

Working with the Banking sample

This example demonstrates how to explore and work with a pre-built sample:

Step	Action	Prompt
1	List the available samples	`What samples are available?`
2	Import the sample into a new automation	`Import the Banking sample in a new automation called "Banking Samples"`
3	Explain the decision service	`Explain the Loan Approval decision service`
4	Summarize a decision table	`Summarize the "Grade" decision table`
5	Update the decision table	`Increase the repayment from 30000 to 35000 in the decision table`

Creating a decision service to evaluate car prices

This example shows how to build a decision service from scratch:

Step	Action	Prompt
1	Create the data and decision models	`Create a data model with the original car price, the age, the brand, the mileage, the list of accidents with their date and severity, then create a decision model to evaluate the car price`
2	Generate test data	`Create 5 test data sets for this model`
3	Run tests	`Run the test data sets`