v2.128.0
26 days ago by Jordan Fanapour
New Features
- Added comprehensive EffortLevel support to AIPromptModel with priority hierarchy system allowing model-specific effort level overrides
- Introduced new Grok 4-1 Fast models (both reasoning and non-reasoning variants) with 2M token context window and optimized pricing
- Added custom Angular component support for artifact viewer tabs with ComponentChangeRequest and ComponentTestResult types
- Implemented advanced dashboard view configuration panel with enhanced filtering and layout options
- Added data requirements viewer component for better artifact visualization and analysis
- Introduced notification resource wrapper for improved Explorer UI navigation
Improvements
- Fixed GPT 5.2 naming convention to align with OpenAI standards (changed from "GPT-5.2" to "GPT 5.2")
- Enhanced Gemini provider with token usage reporting and Flash 3 effort level optimization
- Improved markdown service with post-processing to unwrap incorrectly code-wrapped HTML content
- Enhanced shell component with advanced notification management and view state persistence
- Upgraded artifact viewer panel with better layout controls and component integration
- Improved Explorer core routing with resource wrapper integration for better user experience
Bug Fixes
- Fixed Gemini Flash 3 effort level configuration for optimal performance
- Suppressed mermaid visual error rendering in markdown component to prevent display issues
- Resolved view selector styling issues and improved responsive design
- Fixed notification service integration issues in Explorer shell component
- Corrected AI prompt runner effort level precedence hierarchy implementation
- Updated package dependencies across 143 packages for better compatibility and security