…ects/2.2.0

#46536)

* Fix pylint complaining about too many if-else branch in get_openai_client implementation

* Resolved comments

* Remove user-agent construction tests for sync and async clients

* Add telemetry models and update API view properties for enhanced telemetry support

* Enhance sample_session_log_stream.py to verify and assign Azure AI User role for hosted agent identity

* rbac

* Refactor hosted agent samples to streamline session management and RBAC setup

* update rbac util

* tsp loc

* update hosted agent samples

* update recording

* refactor test hosted agents samples to handle RBAC complications

* change log

* samples for async

* Update installation instructions to include aiohttp for async samples

…Azure/azure-sdk-for-python into feature/azure-ai-projects/2.2.0

* Fixed telemetry instrumentor to correctly call is_recording() as a method

* adding tests

* updating changelog

…Azure/azure-sdk-for-python into feature/azure-ai-projects/2.2.0

…s.cmd

…Azure/azure-sdk-for-python into feature/azure-ai-projects/2.2.0

…ects/2.2.0

…ects/2.2.0

* Enhance sample script for Azure AI evaluations

Updated the sample script to support additional command-line arguments for trace evaluations, including agent ID and trace IDs. Modified the lookback hours default value and improved the overall structure for better clarity.

* Add sample for Azure AI Evaluations with hosted agent

This sample demonstrates how to run Azure AI Evaluations against a hosted agent using the azure_ai_target_completions data source, evaluating agents live with built-in quality and safety evaluators.

* Revert TRACE_LOOKBACK_HOURS default to 1 hour to match docstring

Agent-Logs-Url: https://github.com/Azure/azure-sdk-for-python/sessions/f15ae794-b773-4e8c-860b-5aea55873600

Co-authored-by: shrutiyer <9905402+shrutiyer@users.noreply.github.com>

* Add agent eval samples: trace evals and agent-as-target evals

Add two new evaluation samples:

- sample_evaluations_builtin_with_traces.py: Trace-based evaluation with
  three modes (client-side App Insights query, server-side agent ID, and
  explicit trace IDs)
- sample_evaluations_agent_as_target.py: Live agent evaluation using
  azure_ai_target_completions data source

Both samples use the azure_ai_evaluator config pattern with builtin
intent_resolution and task_adherence evaluators.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Refactor evaluator configuration functions

* Delete sdk/evaluation/azure-ai-evaluation/samples/sample_evaluations_agent_as_target.py

* Delete sdk/evaluation/azure-ai-evaluation/samples/sample_evaluations_builtin_with_traces.py

* Add sample_evaluations_agent_as_target to skip list

* Fix pyright error in trace evaluation samples

Use list() to narrow args.trace_ids from Optional[List[str]] to List[str]
inside the elif guard, resolving pyright reportArgumentType error.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Skip sample_evaluations_agent_as_target in test discovery

The sample requires a deployed Azure AI Agent and has no recording,
so it must be excluded from the parametrized test_evaluation_samples.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Refactor the sample

* pylint fixes

* more pylint fixes

* rename

* rename

* one leftover rename

* undo naming

* fix other samples pylint

---------

Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: Shruti Iyer <you@example.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

…6667)

Rename AgentEndpoint to AgentEndpointConfig per TypeSpec commit f6fde1c.
Updated samples to use the new model name.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

…, remove isolation_key) (#46704)

…ects/2.2.0

…ons) (#46776)