Add OpenAI function calling support

.env.example

@@ -11,4 +11,13 @@ PORT=8000
 MAX_TIMEOUT=600000
 
 # CORS Configuration
-CORS_ORIGINS=["*"]
+CORS_ORIGINS=["*"]
+
+# Rate Limiting Configuration
+RATE_LIMIT_ENABLED=true
+RATE_LIMIT_PER_MINUTE=30
+RATE_LIMIT_CHAT_PER_MINUTE=10
+RATE_LIMIT_DEBUG_PER_MINUTE=2
+RATE_LIMIT_AUTH_PER_MINUTE=10
+RATE_LIMIT_SESSION_PER_MINUTE=15
+RATE_LIMIT_HEALTH_PER_MINUTE=30

CHANGELOG.md

@@ -0,0 +1,43 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+## [1.1.0] - 2025-07-06
+
+### 🎉 Major Features
+
+- **OpenAI Function Calling Support** - Full implementation of OpenAI's function calling format
+  - Complete compatibility with OpenAI's `tools` and `tool_choice` parameters
+  - Support for legacy `functions` and `function_call` format
+  - All 9 Claude tools available through OpenAI-compatible interface
+  - Tool response handling with proper `tool_calls` format
+  - GET /v1/tools endpoint to list available tools
+
+### 🔧 Technical Improvements
+
+- Added comprehensive tool mapping system (OpenAI names → Claude tools)
+- Implemented tool execution handler with proper error handling
+- Enhanced Swagger/OpenAPI documentation with tool schemas
+- Added production-ready test suite for tool functionality
+- Improved message handling for tool responses
+
+### 📚 Documentation
+
+- Updated README with complete function calling examples
+- Added tool usage documentation with all three supported formats
+- Created comprehensive examples in `examples/tools_example.py`
+- Enhanced API documentation with tool-related endpoints
+
+### 🧪 Testing
+
+- Created extensive test suite for tool functionality
+- Added actual execution demonstrations
+- Implemented production readiness checks
+- Verified all tool mappings and structures
+
+## [1.0.0] - Previous Release
+
+- Initial release with core OpenAI compatibility
+- Session management and continuity
+- Multi-provider authentication support
+- Streaming and non-streaming responses

MERGE_SUMMARY.md

@@ -0,0 +1,87 @@
+# Merge Summary: Security & Performance Improvements
+
+## Successfully Merged from Upstream
+
+### 1. **Security Enhancements** ✅
+- **CVE Fix**: Updated `python-multipart` from 0.0.12 to 0.0.18 to resolve critical security vulnerabilities
+- **Rate Limiting**: Implemented comprehensive rate limiting for all endpoints
+  - Chat completions: 10 requests/minute
+  - Debug endpoint: 2 requests/minute  
+  - Auth status: 10 requests/minute
+  - Health check: 30 requests/minute
+  - Sessions: 15 requests/minute
+- **API Key Verification**: Added authentication check to `/v1/models` endpoint
+
+### 2. **New Features** ✅
+- `rate_limiter.py`: Complete rate limiting implementation using SlowAPI
+- Configurable via environment variables
+- JSON error responses with retry-after headers
+- Per-endpoint customizable limits
+
+### 3. **Updated Dependencies** ✅
+- `python-multipart`: ^0.0.12 → ^0.0.18 (security fix)
+- `slowapi`: ^0.1.9 (new dependency for rate limiting)
+
+## Your Existing Features Preserved
+
+All your enhancements remain intact:
+- ✅ OpenAI function calling support
+- ✅ Swagger UI integration  
+- ✅ Enhanced session management
+- ✅ Tool handler and registry
+- ✅ Enhanced parameter validation
+
+## Testing Recommendations
+
+1. **Test Rate Limiting**:
+   ```bash
+   # Test rate limit on chat endpoint
+   for i in {1..15}; do
+     curl -X POST http://localhost:8000/v1/chat/completions \
+       -H "Content-Type: application/json" \
+       -d '{"model": "claude-3-5-sonnet-20241022", "messages": [{"role": "user", "content": "Hi"}]}'
+   done
+   ```
+
+2. **Verify Function Calling Still Works**:
+   ```bash
+   # Test with your existing function calling code
+   python test_tools.py
+   ```
+
+3. **Check Swagger UI**:
+   - Visit http://localhost:8000/docs
+   - Ensure all endpoints are documented
+
+4. **Test Security**:
+   - Verify API key protection works if configured
+   - Check rate limiting responses return proper JSON
+
+## Environment Variables
+
+Add these to your `.env` file:
+```bash
+# Rate Limiting
+RATE_LIMIT_ENABLED=true
+RATE_LIMIT_CHAT_PER_MINUTE=10
+RATE_LIMIT_DEBUG_PER_MINUTE=2
+RATE_LIMIT_AUTH_PER_MINUTE=10
+RATE_LIMIT_SESSION_PER_MINUTE=15
+RATE_LIMIT_HEALTH_PER_MINUTE=30
+```
+
+## Next Steps
+
+1. Install new dependencies:
+   ```bash
+   pip install slowapi
+   ```
+
+2. Test the merged features thoroughly
+
+3. Push to your branch:
+   ```bash
+   git push origin merge-upstream-improvements
+   ```
+
+4. Create a pull request to review changes before merging to main/production

MISSING_UPSTREAM_FEATURES.md

@@ -0,0 +1,96 @@
+# Missing Upstream Features & Improvements
+
+## Features Not Yet Merged from Original Repository
+
+### 1. **GitHub Actions Workflows** 🤖
+The upstream has two GitHub Actions for automated code review:
+
+- **`.github/workflows/claude-code-review.yml`**: Automated PR code reviews using Claude
+- **`.github/workflows/claude.yml`**: Claude PR assistant workflow
+
+These provide:
+- Automatic code review on pull requests
+- AI-powered suggestions and improvements
+- Automated security checks
+
+### 2. **Docker Improvements** 🐳
+While you have your own Docker setup, the upstream has:
+
+- **Standard `docker-compose.yml`**: Simpler compose file for basic deployment
+- **Different Dockerfile approach**: Their Dockerfile might have optimizations
+
+Your setup appears more advanced with:
+- `docker-compose.dev.yml` for development
+- `deploy-dev.sh` and `deploy-prod.sh` scripts
+- Your own Dockerfile
+
+### 3. **Startup Optimization** ⚡
+Commit `8af376a`: Uses Claude 3.5 Haiku for faster/cheaper startup verification
+```python
+# In claude_cli.py - uses Haiku model for verification
+model="claude-3-5-haiku-20241022"  # Faster and cheaper
+```
+
+### 4. **Documentation Updates** 📚
+Several README improvements for:
+- Docker deployment instructions
+- Performance optimization tips
+- Updated examples using Haiku model
+
+## Your Unique Features (Not in Upstream)
+
+You have many features the upstream doesn't have:
+
+1. **OpenAI Function Calling** ✅
+2. **Swagger UI** (`openapi.yaml`) ✅
+3. **Advanced Tool System** (`tool_handler.py`, `tools.py`) ✅
+4. **Production Deployment Scripts** ✅
+5. **Development Docker Compose** ✅
+6. **Extensive Testing Suite** ✅
+7. **Session Management Enhancements** ✅
+8. **Parameter Validation System** ✅
+
+## Recommendations
+
+### Worth Cherry-Picking:
+1. **Startup Optimization** - Easy win for faster startup:
+   ```bash
+   git cherry-pick 8af376a
+   ```
+
+2. **GitHub Actions** - If you want automated PR reviews:
+   ```bash
+   git checkout upstream/main -- .github/workflows/
+   ```
+
+### Already Have Better Versions:
+- **Docker Setup**: Your setup with dev/prod scripts is more sophisticated
+- **Documentation**: You have your own comprehensive docs
+
+### Optional Considerations:
+- Review their Dockerfile for any optimizations
+- Check if their docker-compose.yml has useful environment variables
+
+## Quick Command to Get GitHub Actions
+
+If you want the automated code review features:
+
+```bash
+# Create .github directory and copy workflows
+mkdir -p .github/workflows
+git checkout upstream/main -- .github/workflows/claude-code-review.yml
+git checkout upstream/main -- .github/workflows/claude.yml
+
+# Commit the changes
+git add .github/
+git commit -m "Add Claude Code GitHub Actions for automated PR reviews"
+```
+
+## Summary
+
+You're only missing:
+1. GitHub Actions for automated reviews (optional)
+2. Startup optimization using Haiku model (recommended)
+3. Some documentation updates (low priority)
+
+Your fork is actually MORE feature-rich than the upstream in most areas!