Agentic azd - benchmarking and evaluation

[JeffreyCA]

We could start by measuring how well the agentic init is able to successfully initialize and deploy a curated set of sample apps/templates to Azure. This would allow us to quantify and track the improvements made to the agent.

[spboyer]

An evaluation framework now exists at cli/azd/test/eval/ with test graders and validators. PR #7202 further expands this with a comprehensive evaluation and testing framework.

Closing as implemented.