v1.112.1

✨ Enhanced Script Expansion
Improved trace details for script expansion by adding an expanded: true property, offering better debugging insights.

⚡ DeepSeek V3 and R1 Model Support
Added configurations for DeepSeek V3 and R1 models across multiple providers, including Azure AI Inference. Expanded model options in ModelType.

📜 Versioned Type Definitions
Generated .d.ts files now include the package version for better traceability and consistency.

🔍 Markdown-Enhanced Descriptions
Introduced markdown support in JSON schema descriptions, enabling richer and more readable documentation.

🔧 Refined Search and Debugging

• Enhanced grep functionality in docs search with better query handling.
• Added debug traces for tool definitions in JSON format for improved troubleshooting.

🎨 UI Improvements

• Script groups (excluding system) now open by default in the VS Code extension.
• Collapsible elements for prompt parameters refined for a cleaner interface.

🐛 Bug Fixes and Tweaks

• Fixed issues with GPT-4 encoder tests.
• Addressed collapsible UI element behavior in the web app.
• Adjusted token limits and logging in sample scripts for better usability.

v1.112.0

What's New in GenAIScript 1.112.0

🚀 Enhanced Model Flexibility

• Added support for expanded LLM configurations, including reusable aliases for model mappings.
• Introduced a new configuration file for streamlined model management.

🛠️ Improved Developer Experience

• Enhanced interactive Git commit workflows with new options and reusable push functionality.
• Upgraded CLI with better script filtering, expanded vector search capabilities, and refined embeddings handling.

🔍 Smarter Search Tools

• Added minimum score filtering for vector and fuzzy search, improving relevance in retrieval workflows.
• Enhanced RAG (Retrieval-Augmented Generation) support with optimized document chunking and search precision.

📖 Documentation Updates

• Included new YouTube links in the README, featuring "Mr. Maeda's Cozy AI Kitchen" for an engaging learning experience.

✨ Miscellaneous Enhancements

• Improved error handling and logging across various modules.
• Added new sample scripts for C++ repair and RAG workflows.

Upgrade now to explore these exciting features and improvements!

v1.111.4

What's New in 1.111.4 🎉

• Enhanced LLM Documentation: Updated the LLM-friendly docs URL to provide full markdown content for seamless integration into RAG systems.
• UI Schema Improvements: Introduced uiSuggestions for string fields in JSON schemas, allowing better suggestion handling in forms.
• Refactored JSON Schema Forms: Modularized and optimized form components for cleaner code and improved maintainability.
• Sample Config Update: Added uiSuggestions example in the sample parameters file to demonstrate new functionality.

Enjoy a smoother, smarter GenAIScript experience! 🚀

v1.111.3

✨ Enhanced Model Handling

• Introduced support for F* files with improved MIME type recognition.
• Enhanced model configuration options, including new constants and better alias resolution.

🛠️ CLI Improvements

• Added --format option to the models list command, enabling output in JSON or YAML.
• Improved environment and model information display with refined configurations.

📚 LLM-Friendly Documentation

• Streamlined access to markdown-rendered docs for seamless integration with Retrieval-Augmented Generation (RAG) systems.

🎨 Sample Updates

• Updated sample scripts to include new model configurations and F* file support.

v1.111.2

✨ Enhanced Model Resolution Logic
Refined the criteria for determining model alias candidates, improving connection accuracy and reliability.

🎨 Improved Core Module Functionality
Upgraded unfencing logic to support multiple languages, enhanced utility functions like arrayify, and streamlined imports for better maintainability.

✨ New Parser Features
Introduced a dedent function for removing left indentation and expanded parser capabilities with enhanced testing coverage.

📜 New Prompt Template
Added a demo "joke-template" for categorizing jokes as funny or not, showcasing interactive prompt features.

🖼️ Richer Sample Outputs
Updated sample outputs with new detailsFenced sections and refined test data for better clarity and usability.

v1.111.1

🎉 Enhanced Traceability
Tool render outputs in terminal logs now include call IDs, making debugging and tracking easier.

🐛 Improved Error Handling
API responses now return clearer error messages for better troubleshooting.

🛠️ Flexible GitIgnore Handling
Introduced options to enable or disable .gitignore filtering, offering more control over file processing.

🖼️ New Image-to-Prompt Script
Added a sample script for generating prompts based on tiled image assets, tailored for DALL-E-3.

📂 File Expansion Improvements
Refined file resolution logic to respect ignoreGitIgnore settings and handle exclusions more effectively.

v1.111.0

✨ Enhanced Model Support

• Improved reasoning configurations and added new aliases for more flexible model handling.
• Expanded support for Anthropic models, including enhanced reasoning effort capabilities and extended output options.

🖼️ Image Generation API

• Introduced a basic image generation API, enabling creation of visuals directly from prompts.

🔧 DevContainer Setup

• Added a DevContainer configuration for streamlined development in VS Code environments.

🖍️ Console Improvements

• Revamped terminal message rendering with enhanced formatting and color support.

🚀 Dependency Updates

• Upgraded TypeScript and other dependencies to their latest versions for improved performance and security.

🛠️ Azure OpenAI Enhancements

• Added image generation, transcription, and speech capabilities for Azure OpenAI integration.

🎨 UI Threshold Updates

• Adjusted control chat thresholds to improve user experience in the interface.

v1.110.3

✨ Enhanced JSON Schema Validation
Improved schema validation with added support for arrays and stricter property inference. Tests now cover more edge cases, ensuring robust validation.

🆕 Optional Integer Parameter
Introduced an optional integer parameter in the configuration, offering more flexibility for user-defined scripts.

📂 Expanded File Support
Added support for *.prompty files in the accept parameter, alongside existing formats like .txt, .md, and .pdf.

🐛 JSONSchemaNumber Improvements
Refactored the JSONSchemaNumber component to explicitly handle the required property, enhancing clarity and functionality.

🔧 Code Cleanup
Streamlined parsing logic for file extensions and removed unused utility functions for better maintainability.

v1.110.2

✨ Enhanced Performance Tracking
Introducing a new performance monitoring system with detailed logging and measurement capabilities to boost efficiency and identify bottlenecks.

🚀 Smarter File Handling
Replaced readFile with a robust tryReadText function for safer and more reliable file operations, ensuring smoother error handling.

📊 Model Management Improvements
Added a models command for better visibility and control of language model alias mappings, enhancing user interaction with model configurations.

🐛 Error Logging Fixes
Ensured error directories are created before logging, preventing potential disruptions during error reporting.

⚡ Optimized Token Estimation
Implemented a faster, approximate token estimation method for improved performance in token-heavy operations.

🛠️ Refactored Components
Streamlined configure.ts and other key modules for better code clarity and maintainability.

v1.110.1

🚀 Enhanced DOCX Parsing

• Improved DOCX file parsing with robust error handling and caching for faster processing.
• Added support for multiple output formats: Markdown, HTML, and plain text.

🐞 Bug Fixes

• Resolved issues with file parsing (#1205, #1198).

🔧 Developer Experience

• Introduced new test cases for DOCX parsing to ensure reliability.
• Enhanced form fields in the web interface with better input handling and auto-completion disabled.

📏 Hash Length Update

• Extended hash lengths for DOCX and PDF processing to improve uniqueness and cache accuracy.