🗣️ DebateGraph: Multi-Agent AI Debate System

DebateGraph is an intelligent Multi-Agent System (MAS) that simulates structured debates between opposing AI personas. Built using LangGraph and LangChain, it orchestrates a strictly controlled workflow where agents maintain memory, adhere to turn-based constraints, and submit to a final verdict by an AI Judge.

---

🚀 Features

• Dual-Agent Architecture: Simulates a clash between a Scientist (Empirical/Logic-driven) and a Philosopher (Ethics/Value-driven).
• Structured Workflow: Enforces a strict 8-round debate format using a Directed Acyclic Graph (DAG).
• State Management: Uses LangGraph to manage conversation state, turn history, and "next speaker" logic.
• Smart Memory: Agents receive "sliced" context (only the relevant previous argument) to prevent context bloating.
• AI Judge: A specialized node analyzes the full transcript to declare a winner based on logical coherence and persuasion.
• High-Performance Inference: Optimized for Groq API (Llama 3.3) for lightning-fast responses without rate-limit bottlenecks.
• Audit Logging: Automatically generates timestamped JSON logs for every state transition and argument.

---

🛠️ Tech Stack

• Orchestration: LangGraph (StateGraph)
• LLM Framework: LangChain Core
• Inference Engine: Groq API (Llama-3.3-70b-versatile)
• Language: Python 3.10+
• Utilities: Pydantic (Validation), Python-Dotenv (Config)

---

⚙️ How It Works (The Graph)

The system operates as a State Machine with the following nodes:

1. UserInput: Initializes the debate topic.
2. Controller (Edge): Checks round count. If < 8, routes to the next speaker; if 8, routes to Judge.
3. Agent Nodes (A/B): Generates arguments using specific system prompts (Scientist vs. Philosopher).
4. Memory Node: Updates the conversation history and flips the "turn" flag.
5. Judge Node: Critical analysis of the full debate logic and declares a winner.

(Note: Diagram is generated automatically on the first run)

---

📦 Installation

1. Clone the Repository

   git clone [https://github.com/Nikhil-Codehub/DebateGraph.git](https://github.com/Nikhil-Codehub/DebateGraph.git)
   cd DebateGraph

2. Create a Virtual Environment

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # Mac/Linux
   source venv/bin/activate

3. Install Dependencies

   pip install -r requirements.txt

4. Set up Environment Variables Create a .env file in the root directory:

   GROQ_API_KEY=gsk_your_groq_api_key_here

---

🏃 Usage

Run the main script via CLI:

python main.py