- Implemented a new method in the CharacterCreatorMixin to extract or generate dialogue examples for characters based on provided text.
- Updated the character card loading process to include dialogue example determination, ensuring examples are regenerated properly.
- Created a new Jinja2 template for generating dialogue examples, including guidelines for format and content.
- Enhanced logging for dialogue example generation to track character names and example counts.
- Created a new template to generate prompts for horizontal illustrations capturing dynamic moments in scenes.
- Included sections for character context, requirements, and task instructions to guide image generation.
- Emphasized the importance of action, emotion, and cinematic framing in the generated images.
- Introduced a require_active boolean flag in various reinforcement classes and templates to control reinforcement activation based on character status.
- Updated logic in WorldStateAgent to skip inactive character reinforcements when require_active is true.
- Enhanced frontend components to support the new require_active option for character reinforcements.
- Added a "Re-analyze" button that appears when file data is present and analysis is not in progress.
- Improved error handling by introducing a "Retry" button in the error alert for failed analyses.
- Adjusted the display logic for analysis information to improve user experience.
- Expanded the list of excluded names in character detection to include additional variations, preventing false positives during character name identification.
- Introduced a new method, detect_characters_from_texts, to analyze multiple texts by processing them in manageable chunks based on the client's max context size.
- Added functionality to avoid duplicate detections by passing already detected character names.
- Implemented utility functions for chunking items by token count and removing substring names to improve character detection accuracy.
- Updated the corresponding Jinja2 template to reflect the changes in character detection logic.
