Python: Fix runtime response format for responses client

[moonbox3]

Motivation and Context

Previous work was done #2362 to allow for passing in a JSON schema to be used as a response format for OpenAI chat clients. However, when using this for the Responses client, it is breaking.

The PR brings in some changes to normalize response_format into Responses text.format and route Pydantic models through parse only when needed. Guard against conflicting response_format/text combinations and reuse mapping for streaming responses.

Description

• Fix runtime schema for Responses client.
• Add sample in getting started section.

Contribution Checklist

• The code builds clean without any errors or warnings
• The PR follows the Contribution Guidelines
• All unit tests pass, and I have added new tests where possible
• Is this a breaking change? If yes, add "[BREAKING]" prefix to the title of the PR.

[markwallace-microsoft]

Python Test Coverage Report •

File	Stmts	Miss	Cover	Missing
packages/azure-ai/agent_framework_azure_ai
_client.py	160	28	82%	167–170, 174–175, 177, 218, 220–223, 225, 355–356, 360, 362–363, 365–373, 375
packages/core/agent_framework/openai
_responses_client.py	460	90	80%	98, 144, 151–152, 155–156, 161–162, 165, 172, 191, 198, 201–208, 210, 215, 221, 224, 239, 242–243, 245, 272, 302, 330–331, 358, 362, 379, 384, 426, 490, 565, 570, 574–576, 596, 611–612, 616–618, 666, 686–687, 700–701, 717–718, 756, 758, 796, 798, 807–808, 821, 823, 896–902, 919–924, 943, 961, 971, 973, 991–992, 994–996, 1007–1008, 1011, 1013
TOTAL	15538	2378	84%

Python Unit Test Overview

Tests	Skipped	Failures	Errors	Time
2227	130 💤	0 ❌	0 🔥	54.193s ⏱️

[Copilot]

Pull request overview

This PR fixes runtime response format handling for the OpenAI Responses client by normalizing response_format into the Responses text.format configuration. The changes enable passing JSON schemas at runtime via additional_chat_options without defining Pydantic models upfront.

Key changes:

• Added helper methods _prepare_text_config and _convert_response_format to normalize response format configurations
• Modified both streaming and non-streaming response methods to handle runtime JSON schemas
• Added a new sample demonstrating runtime JSON schema usage with the Responses client

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated no comments.

File	Description
python/packages/core/agent_framework/openai/_responses_client.py	Core implementation: adds logic to normalize response_format into text configuration, routing Pydantic models through parse and runtime schemas through text.format
python/samples/getting_started/agents/openai/openai_responses_client_with_runtime_json_schema.py	New sample demonstrating runtime JSON schema usage with both streaming and non-streaming responses
python/samples/getting_started/agents/openai/openai_chat_client_with_runtime_json_schema.py	Added type annotation to chunks variable
python/samples/getting_started/agents/openai/README.md	Added documentation entry for the new runtime JSON schema sample

---

You can also share your feedback on Copilot code review for a chance to win a $100 gift card. Take the survey.