In Part 1 of this blog series set the stage for why platform engineering is being reshaped by agentic AI. (read it here)

Basically we outline how instead of humans translating intent through layers of CLIs, SDKs, and bespoke workflows, capable agents can interpret natural-language goals and turn them into safe, validated platform actions using well-described APIs and control schemas.

That shift changes what “good” looks like for internal platforms, raising the bar on guardrails, policy, and self-service interfaces allowing teams to move faster without sacrificing safety, reliability, or governance.

In this article (Part 2), we’ll build on those principles with a practical example that shows how you can design and operate a platform for this new era.

Let’s Go APE!

1. Getting Started

To get started with APE’s Git-ape, clone the GitHub repository to your local machine. This gives you a complete copy of the project so you can explore the code, run it locally, or start contributing.

Open a terminal and run:

git clone https://github.com/Azure/git-ape
cd git-ape

You’re now ready to begin working with Git-ape locally.

Git-ape provides a ready‑to‑use demo experience so you can quickly explore how Git-ape generates and manages cloud infrastructure using Infrastructure‑as‑Code.

2. Prerequisites

Before you begin, make sure you have the following installed and configured:

• Visual Studio Code with the GitHub Copilot extension enabled
• Only tested with BASH shells Ubuntu & Git-Bash on windows
• Azure CLI installed and authenticated

1. az login

• Azure MCP Server configured (see below for example)

3. Installation

This repository contains agent configuration files that GitHub Copilot automatically discovers.

• Open the workspace in VS Code
• Verify the agents are available by typing:
  • @git-ape for deployments
  • @Git-ape Onboarding for repository setup

If the agents appear, they are correctly installed and ready to use.

MCP Server Configuration

To enable Git-ape’s Azure capabilities, configure Azure MCP in your VS Code settings.

Open your VS Code settings.json and add:

{
"azureMcp.serverMode": "namespace",
"azureMcp.enabledServices": [
"deploy",
"bestpractices",
"group",
"subscription",
"functionapp",
"storage",
"sql",
"monitor"
],
"azureMcp.readOnly": false
}

This configuration enables Git-ape to deploy resources, apply best practices, and interact with Azure services directly through Copilot.

For advanced scenarios and troubleshooting, see the Azure MCP Setup Guide.

3. Destruction

To remove your deployed environment use the @git-ape to clean up after you.

@git-ape destroy [your mess]

You’re Ready to APE about!

With the repository cloned, agents discovered, and Azure MCP configured, you’re ready to use Git-ape through GitHub Copilot to deploy and onboard Azure resources from within VS Code. Use the ReadME to build and deploy the demo examples.

This video walk through some of these examples to give you a taste of the experience: