feat: Update proxy implementation with Azure OpenAI support from PR #35

[rysweet]

Feature Request: Azure OpenAI Proxy Integration

Objective

Update the existing proxy implementation to integrate Azure OpenAI support from fuergaosi233/claude-code-proxy#35 while maintaining full backward compatibility with current OpenAI configurations.

Requirements

Functional Requirements:

• Integrate code changes from claude-code-proxy PR Bring codebase to pyright type safety compliance #35
• Add Azure OpenAI service support alongside existing OpenAI support
• Support new reasoning models (GPT-5, O3, O4) with proper model mapping
• Handle max_completion_tokens parameter for token limit control
• Maintain all existing AmplIHack proxy integration points

Non-Functional Requirements:

• Zero breaking changes for existing OpenAI users
• Performance parity with current implementation
• Secure credential handling for Azure configurations
• Comprehensive error handling and logging

User Story

As a developer using AmplIHack
I want to configure Azure OpenAI as my LLM provider
So that I can leverage Azure's enterprise features while keeping my existing workflow

Technical Considerations

Architecture Impacts:

• Extend ProxyConfig class for Azure-specific configuration parameters
• Update ProxyManager to handle Azure OpenAI client initialization
• Add model mapping logic for Azure deployment names
• Integrate reasoning model support from PR Bring codebase to pyright type safety compliance #35

Dependencies:

• claude-code-proxy PR Bring codebase to pyright type safety compliance #35 changes
• Azure OpenAI SDK compatibility
• Existing AmplIHack proxy integration layer

Integration Points:

• Current proxy configuration system
• Model selection and routing logic
• Authentication and credential management
• Error handling and retry mechanisms

Acceptance Criteria

• All existing OpenAI configurations continue to work without changes
• Azure OpenAI configurations can be set up and function correctly
• Model mapping works for both OpenAI and Azure deployment names
• Reasoning models (GPT-5, O3, O4) are supported on both providers
• max_completion_tokens parameter is handled properly
• All existing tests pass
• New tests added for Azure OpenAI functionality
• Documentation updated with Azure configuration examples
• No performance regression in existing functionality
• Proper error messages for misconfigured Azure settings

Implementation Notes

Key Files to Modify:

• Proxy configuration classes
• ProxyManager initialization logic
• Model mapping and routing functions
• Integration with AmplIHack's proxy system

Reference Implementation:
fuergaosi233/claude-code-proxy#35

Complexity Assessment

Complexity: Medium

• Multiple configuration paths to support
• External dependency integration required
• Backward compatibility constraints
• Model mapping logic complexity

Estimated Effort: 2-3 days

Priority: High

This feature enables enterprise Azure OpenAI usage while maintaining the existing OpenAI workflow, expanding AmplIHack's deployment options significantly.

🤖 Generated with Claude Code