AI Rule Authoring from GitHub Copilot CLI

AI Knowledge Share
1

Pega’s Infinity Studio has introduced fantastic generative AI capabilities for conversational rule authoring within the portal. But what if you are a professional front-end developer building custom Constellation DX components or custom Front Ends?

Constantly bouncing back and forth between your IDE (where you write React or Angular code) and Dev Studio just to configure a UI view ruins your development flow.

By utilizing new Infinity AI Plugins (via the pegasystems/infinity-ai-plugins repository), you can now author Pega rules and update views directly from your local coding workspace.

In this demo, we bypass the Pega UI entirely to author and configure rules directly from a local terminal:

1. Zero-Friction Setup

Install the plugin into copilot CLI using the instructions given in pegasystems/infinity-ai-plugins repository. Restart the session for the plugin and the MCP to be loaded.

Using the copilot CLI, a quick command pega-setup establishes our secure connection to the Pega Infinity runtime. It asks for login for the first time after setup. The agent checks our environment, hooks into the MCP (Model Context Protocol) server, and points to default application of the logged in user.

2. Conversational Authoring

Instead of manually switching the application, finding the correct view, and dragging fields in, we simply instruct our agent to do those tasks and it performs them seamlessly. Here we take a simple example of creating a property and adding it to a view.

3. Under the Hood: Hands-free Integration

The local agent immediately utilizes Pega’s API to:

  • Create a dedicated development branch (AddUserNameFld) to isolate changes.

  • Generate the new UserName text property rule.

  • Update the structural layout of the Create view rule to include our property.

4. The Result

When we open the branch in Infinity Studio, we can see a branch created and contains the new property and view in it. Also, when we open UI Designer and create a new case, we can see our new field appearing in the create view.

Why This Matters for Developers using IDEs?

If you are developing proprietary UI components or custom frontends, this workflow bridges the gap between raw code and low code. You can make changes to your custom component (like adding new input props), and in the same terminal session, publish it to Pega and command your agent to make changes in the view rule to assign respective values to these input props.

It keeps the developers in their preferred environment and automates the platform configuration.

6 Likes
2

To author and configure Pega rules without leaving your local coding workspace, install the Infinity AI Plugins from the pegasystems/infinity-ai-plugins repository into your GitHub Copilot CLI and run the pega-setup command to securely connect to your Pega Infinity runtime. Once authenticated, you can conversationalize your workflow by using the terminal agent to automatically create development branches, generate properties, and update UI views through Pega’s API. This integration eliminates the need to constantly switch back and forth between your local IDE and the Pega Dev Studio portal. By remaining in your terminal, you can seamlessly publish custom front-end components and direct the AI agent to update the corresponding Pega view rules instantly.