Run a minimal Socratic seminar pipeline locally.
Prerequisites:
- Python 3.10+
- OpenAI API key in environment:
OPENAI_API_KEY - Redis server running locally (e.g., via Docker)
- Ensure Redis is running before starting services; the broker will attempt to reconnect if it starts first.
-
Start Redis: If you have Docker installed, start Redis in the background by running this command in your terminal:
docker compose up -d -
Install dependencies:
pip install -r requirements.txt
-
Set up your environment:
- Create a
.envfile in the project root - Add your OpenAI API key:
OPENAI_API_KEY=your_key_here
- Create a
-
Start all services:
python run_all.py
This will start:
- Visualizer (web interface)
- Planner
- Moderator
- Scribe
- Meta Agent
- Multiple Philosopher agents with different perspectives
- All agents consume type-specific task queues (e.g.,
tasks:planner,tasks:start_round,tasks:speak) and events are fanned out per-consumer (events:scribe,events:meta).
-
Open http://localhost:5000 in your browser (opens automatically)
- Enter your philosophical question
- Watch the discussion unfold in real-time
-
Press Ctrl+C in the terminal to stop all services
If you prefer to run services individually:
-
Start services in separate terminals:
python visualizer/visualizer.py python planner/planner.py python moderator/moderator.py python scribe/scribe.py python meta/meta_agent.py python philosopher/philosopher.py # run multiple with different PHIL_PROFILE env -
Start a session using curl:
curl -X POST -H "Content-Type: application/json" \ -d '{"query":"Is free will compatible with determinism?"}' \ http://localhost:5000/api/submit
-
Open http://localhost:5000/session/ to view
See the "Socratic Arena — MVP Implementation Guide" (MVP.md) for detailed architecture, data schemas, and agent prompts.