Fix: Ensure Goblin monster parsing works by using structured data; ensure Chain Mail equipment parsing from PDF is robust

Enhance Equipment PDF chunking with specific item extraction (e.g. Chain Mail) and OCR cleaning

Add full DM Guide PDF loader to initialize_rag.py

Switch Equipment loader to use Player's Handbook PDF instead of text file

Add Equipment loader to initialize_rag.py and fix Equipment path patching

Aggressively patch settings for SpellParser to fix path issues on HF

Update RAG initialization to load full dataset (Spells, Monsters, Classes, Races, Items) on all environments

Move RAG initialization from Docker build to app startup

Fix initialize_rag.py to use current data structure

Add initialize_rag.py to fix HuggingFace build error

refactor: Organize scripts and notebooks into proper directories

fix: Fix class parser to find section headers not table entries

feat: Add name-weighted retrieval, race extraction, and comprehensive testing

feat: Add D&D RAG system with ChromaDB integration