🚑 Fixes

Updated model type logic in ModelService to detect 'bge' within model names for embedding detection. (#6)

🚀 Improvements

Created centralized PromptTemplateService for maintaining and updating prompts more efficiently. (#5)

Enhanced source citation standards, requiring explicit, standardized citations for document references and web URLs.

Implemented specialized agent personas (decomposer, synthesizer, web analyzer) to boost task specificity and output quality.

Improved context handling with a formatter that prioritizes relevant information, manages truncation effectively, and integrates confidence scores.

Standardized response formatting across various output types for consistency.

Added query intent analysis to better handle query structures and intent before response generation.

Enhanced error handling with clear instructions for insufficient or ambiguous contexts.

Comprehensive documentation added, including a detailed README for prompt systems and best practices.

Initial release