azure/agent-landing-zone ? reverse-engineered prompt

Build me a secure Azure starter kit for an AI agent app that I can deploy in one go.

I want a clean enterprise landing zone that sets up the foundation for a private, zero trust style deployment, with observability and infrastructure as code, and then lets me choose whether to deploy the full app, just the infrastructure, or my own custom agent app on top. Use Microsoft Foundry and the Microsoft Agent Framework approach, and make it feel ready for real company use, not just a demo.

It should include the default agent app pieces, like a UI and an orchestrator, and support optional add ons like ingestion and an MCP server. Please wire it so the deployment is easy to customize, and make sure the docs and defaults guide me toward the right setup for a private enterprise environment. If you need to check current Azure docs or recommended settings online, go ahead and do that.