Version 0.2.0
Changelog
0.2.0 (2025-04-25)
- Major enhancement to component detection for modern tech stacks and architectures
- Added support for monorepo structures with package-specific commit grouping
- Expanded commit type taxonomy with 8 new conventional commit types
- Improved commit message formatting to follow GitHub best practices
- Enhanced non-AI commit generation with more intelligent grouping
- Added sophisticated tech stack detection for 30+ frameworks and libraries
- Updated formatter detection to recognize more linting tools
- Implemented smarter naming for commits with descriptive prefixes
- Improved frontend and backend framework component detection
- Made default timeout configurable through the codebase with a central constant
- Better resource utilization with improved CPU and memory detection
- Fixed cross-platform issues with filenames containing spaces and special characters
- Enhanced Git status parsing using null-terminated output for better handling of complex filenames
- Improved path handling with normalized separators for Windows/Unix compatibility
- Added more robust error handling for Git command failures
- Enhanced logging with detailed filename debugging information
- Added failover mechanism for path normalization when staging files fails
- Implemented strict 50-character limit for commit title lines to comply with GitHub best practices
0.1.9 (2025-04-25)
- Increased default timeout from 10 to 30 seconds to avoid frequent timeouts with larger models
- Fixed issue with truncated filenames when staging files for commit
- Improved error handling for file staging operations with better error messages
- Enhanced git configuration validation with user-friendly messaging
- Added parameter descriptions with defaults in help text
0.1.8 (2025-04-25)
- Improved CLI interface with enhanced progress indicators and loading spinners
- Added automatic download of recommended model if no models are available
- Enhanced model selection UI with model size indicators and recommendations
- Implemented intelligent warm-up for models with realistic prompts
- Added resource-awareness to adapt batch processing based on system capabilities
- Improved parallel processing with automatic fallback for low-resource systems
- Added detailed statistics about system resources during startup
- Fixed spinner display during user interaction for better UX
- Added more informative feedback during model initialization and processing
- Added donation links and sponsor information throughout the application
- Added --version flag to display version and project information
- Optimized memory usage for large repositories with change batching
0.1.7 (2025-04-25)
- Enhanced pre-commit hook detection with automatic retry on failure
- Added automatic hook skipping in non-interactive mode for repositories with pre-commit issues
- Added proper success/failure tracking across multiple commit attempts