Add GPT-5 models support and advanced control parameters

[jamesrochabrun]

Summary

• Added support for GPT-5 model family (gpt-5, gpt-5-mini, gpt-5-nano)
• Introduced reasoning_effort parameter with "minimal", "low", "medium", and "high" options
• Added verbosity parameter to control output token generation

Changes

• Updated chat_completion tool schema to include new parameters
• Enhanced performChatCompletion method to parse and pass reasoning_effort and verbosity
• Updated README with comprehensive documentation of new features
• Aligned with SwiftOpenAI 4.3.2 features from PR #164

Test Plan

• Build package successfully with swift build -c release
• Test with GPT-5 models: "Use gpt-5 to explain recursion"
• Test reasoning_effort: "Ask gpt-5 with minimal reasoning effort to write a SQL query"
• Test verbosity: "Use gpt-5-mini with low verbosity to generate code"
• Test combined parameters: "Use gpt-5 with high reasoning effort and low verbosity to find bugs"

Example Usage

"Use gpt-5 with high reasoning effort to solve this complex problem"
"Generate concise code with gpt-5-mini using low verbosity"
"Ask gpt-5 with minimal reasoning effort for quick instructions"

🤖 Generated with Claude Code