🎲 RuleLawyer

RuleLawyer is a lightweight, extensible framework for building a board game rules assistant using Retrieval-Augmented Generation (RAG). Upload your rulebooks (PDFs), process them into searchable chunks, and query them using a local interface powered by OpenAI.

---

🧠 What It Does

• 🔍 Extracts text from board game rule PDFs
• ✂️ Splits text into manageable chunks
• 🧬 Embeds the chunks using OpenAI's embedding model
• 🤖 Answers questions about rules using semantic search + GPT

---

🚀 Getting Started

1. Clone the repo

git clone https://github.com/vildursn/RuleLawyer.git

2. Install necessary packages

   pip install -r requirements.txt

3. Create a folder inside data/raw for your rulebook, and upload your PDF(s) into that folder

4. Add your OpenAI key to the .env file

5. Run the following

Extract and chunk rulebooks

python src/ingest.py

Embed the text chunks

python src/embed.py

Ask questions (WIP)

python src/gui.py