Hierarchical Multi-Agent Consensus System

A LangGraph-based system that enables three AI agents to debate topics and reach consensus through structured conversation.

Features

• Two-Phase System: Presentation mode (5 iterations) + Consensus mode (until agreement)
• Multiple AI Models: Support for various OpenAI models (GPT-4o, GPT-3.5-turbo, etc.) and Anthropic Claude models
• Web Search Integration: Agents can search the web for information
• JSON Export: Automatic export of all conversation data and statistics
• Modular Design: Easy to modify agent order and personalities
• Structured Output: JSON responses with reasoning and explanations

System Architecture

SUPERVISOR → AGENT_X → AGENT_Y → AGENT_Z → Decision Point
    ↓           ↓         ↓         ↓         ↓
  Start    Continue Debate OR Switch to Consensus Mode
    ↓
  Consensus Loop (until supervisor decides consensus reached)

Setup

1. Install Dependencies:

   pip install -r requirements.txt

2. Configure API Keys:

   • Copy env.template to .env
   • Add your OpenAI and Anthropic API keys

3. Run the System:

   python -m consensus_system.main

File Structure

consensus_system/
├── main.py                 # Main entry point
├── config.py              # Configuration and API keys
├── state.py               # Agent state definitions
├── agents/
│   ├── supervisor.py      # Supervisor agent
│   ├── debater_agents.py # Agent X, Y, Z implementations
│   └── prompts.py         # Prompt templates
├── graph/
│   └── graph_builder.py   # LangGraph structure
├── tools/
│   └── web_search.py      # Web search integration
└── utils/
    └── response_parser.py # Response parsing utilities

Usage

Basic Usage

from consensus_system.main import run_consensus_system

topic = "Should artificial intelligence be regulated by governments?"
result = run_consensus_system(topic)

Customizing Agents

• Modify personalities in agents/prompts.py
• Change agent order in graph/graph_builder.py
• Adjust prompts and behaviors as needed
• Configure different AI models in config.py

Customizing Topics

# In main.py
if __name__ == "__main__":
    topic = "Your custom topic here?"
    result = run_consensus_system(topic)

Configuration

• Max Iterations: Set in state.py (default: 5)
• Models: Configure in config.py
• Consensus Prompt: Modify in config.py

Output

The system provides:

• Real-time console output of agent responses
• Structured state storage for analysis
• Conversation history tracking
• Consensus evaluation results
• JSON export of all data for analysis

JSON Export

After each run, the system exports the following JSON files to a timestamped output folder:

• complete_conversation.json - Full conversation history with metadata
• agent_x_responses.json - All Agent X responses with metadata
• agent_y_responses.json - All Agent Y responses with metadata
• agent_z_responses.json - All Agent Z responses with metadata
• summary_statistics.json - System statistics and counts

Research Applications

This system is designed to study:

• Agent order effects on consensus
• Impact of different AI models
• Prompt engineering effects
• Consensus building dynamics
• Personality effects on debate outcomes
• Topic complexity and consensus difficulty

Requirements

• Python 3.8+
• OpenAI API key
• Anthropic API key
• Internet connection (for web search)