feat: Integrate smoke testing into the pipeline

[Vamshi-Microsoft]

Purpose

This pull request introduces several improvements to the test automation framework and CI/CD workflows, focusing on enhanced RAI (Responsible AI) validation, more robust test reporting, and improved workflow reliability. The main changes include new RAI test utilities and error validation methods, custom pytest reporting features (including step-by-step logs and screenshots on failure), and updates to workflow conditions to better handle job cancellation scenarios.

RAI validation and test automation enhancements

• Added new RAI prompt (rai_prompt) to constants.py, and implemented multiple RAI error message and validation methods in BIABPage (validate_rai_error_message, validate_rai_clarification_error_message, input_rai_prompt_and_send, etc.), supporting more comprehensive RAI scenario testing. [1] [2] [3]
• Introduced a gp marker for Golden Path tests in pytest.ini, and updated test execution logic in test-automation-v2.yml to run only Golden Path tests when appropriate, using the -m gp pytest option. [1] [2] [3] [4]

Test reporting improvements

• Added a custom subtests fixture to conftest.py for step-by-step reporting in the HTML report, and improved failure diagnostics by automatically capturing and linking screenshots for failed tests.
• Enhanced HTML report presentation: changed the report title, renamed the "Duration" column to "Execution Time", and customized test node IDs for better readability (especially for Golden Path tests).

CI/CD workflow reliability

• Updated workflow job conditions in .github/workflows/deploy-orchestrator.yml and .github/workflows/job-deploy.yml to use !cancelled() instead of always(), ensuring jobs do not run if the workflow is cancelled. [1] [2] [3] [4] [5]

Does this introduce a breaking change?

• Yes
• No