docs(giskard-llm): document Azure Foundry v1 via OpenAI

[Hartorn]

Summary

• Document Azure Foundry OpenAI v1 as an OpenAI-compatible endpoint through provider="openai".
• Add mocked coverage that the OpenAI provider accepts /openai/v1/ base URLs without Azure api_version.
• Add routing coverage for chat completions, embeddings, and Responses API using Azure deployment names through one OpenAI-compatible alias.
• Add an azure_foundry_v1 functional CI slice that reuses the existing Azure AI CI credentials and derives the OpenAI v1 base URL from AZURE_AI_ENDPOINT.

Tests

• uv run --extra all pytest tests/test_providers.py tests/test_routing.py -q
• uv run pytest --collect-only tests/functional -m "functional and azure_foundry_v1" -q
• uv run ruff check .
• uv run ruff format --check .

Notes

This is stacked on #2486 because it documents and tests the transport/configuration path added there. It intentionally leaves azure_ai/ behavior unchanged.

[gemini-code-assist]

Code Review

This pull request adds support for Azure Foundry OpenAI v1 endpoints by updating documentation and adding unit tests for routing and provider configuration. The changes ensure these OpenAI-compatible endpoints are correctly handled by the LLMClient. Feedback was provided to improve the formatting of a code example in the README for better consistency.

[gemini-code-assist]

For better readability and consistency with the other calls in this example, consider adding a blank line after this line.

Suggested change

	response = await client.aresponse("foundry-v1/gpt-4.1-mini", "Write one sentence.")
	response = await client.aresponse("foundry-v1/gpt-4.1-mini", "Write one sentence.")