[Epic] Build and operate production AI applications on Azure

## Problem statement

Foundry, AI gateway, hosted agents, models, evaluations, and Copilot SDK workflows are spread across separate capabilities. Developers need a coherent path from application design through deployment and operation.

## Outcome

Developers can build, deploy, evaluate, secure, and operate production AI applications on Azure using connected GitHub Copilot for Azure workflows.

## Goals

- Connect Foundry, model, agent, gateway, evaluation, and application workflows
- Cover production concerns including identity, networking, observability, capacity, and safety
- Provide clear handoffs between development, deployment, and operation

## Non-goals

- Replacing Foundry or Azure MCP product surfaces
- Tracking routine regressions or service incidents

## Success criteria

- [ ] Priority AI application journeys complete end to end
- [ ] Required security and operational checks are included
- [ ] Linked work has measurable acceptance criteria and evaluation coverage

## Dependencies

Azure MCP, Microsoft Foundry, Copilot SDK, evaluation infrastructure, and service owners.

No files, tests, or entry points are named; begin by reviewing the stated Azure MCP, Microsoft Foundry, Copilot SDK, evaluation infrastructure, and service-owner dependencies. Done means priority AI application journeys work end to end with security and operational checks, measurable acceptance criteria, and evaluation coverage.

Written by the indexing model from the issue text.