Add support for Verbosity introduced in GPT-5

[kzu]

See https://platform.openai.com/docs/guides/latest-model?utm_source=chatgpt.com#verbosity

[Copilot]

Pull Request Overview

This pull request adds support for the new Verbosity feature introduced in GPT-5 models, allowing control over the number of output tokens to reduce latency. The implementation follows the same pattern as the existing ReasoningEffort feature.

• Adds a Verbosity enum with Low, Medium, and High levels
• Integrates verbosity support into the OpenAI chat client with proper JSON serialization
• Extends ChatOptions with a Verbosity property for easy configuration

Reviewed Changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 3 comments.

File	Description
src/AI/Verbosity.cs	Defines the new Verbosity enum with Low, Medium, and High options
src/AI/OpenAI/OpenAIChatClient.cs	Implements verbosity support in the OpenAI client with VerbosityOptions class
src/AI/ChatExtensions.cs	Adds Verbosity property to ChatOptions for configuration
src/AI.Tests/OpenAITests.cs	Adds comprehensive test coverage for the verbosity feature

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[kzu]

🧪 Details on Ubuntu 24.04.2 LTS

from dotnet-retest v0.7.2 on .NET 9.0.8 with 💜 by @devlooped