Quick start guide to try the Coagent AI Agent Framework using Docker.
- Docker and Docker Compose
- At least 4GB of available RAM
- Internet connection to download images
-
Clone or download this directory
git clone <repository-url> cd try-coagent
-
Start the demo
docker-compose up
-
Access the application
Once the containers are running, open your browser to:
- Main UI: http://localhost:3000
The demo starts two main services:
- coagent-core (port 3000): Main API server with embedded web UI
Your data is stored locally in the ./coagent_demo_storage
directory.
The docker-compose.yml includes optional services that are commented out:
Uncomment the ollama
service to run local language models. This will:
- Download and run Ollama
- Pull
all-minilm
andqwen3:4b
models - Require significant disk space and memory
# Start the demo
docker-compose up
# Start in background
docker-compose up -d
# View logs
docker-compose logs -f
# Stop the demo
docker-compose down
# Update to latest version
docker-compose pull
docker-compose up
# Use specific version
IMAGE_TAG='v0.4' docker-compose up
- Ensure port 3000 is not in use by other applications
- Check Docker has sufficient resources allocated
- Wait 30-60 seconds after
docker-compose up
for services to fully start - Check logs with
docker-compose logs coagent-core
docker-compose down
rm -rf coagent_demo_storage
docker-compose up
For issues or questions:
- Contact us via the
Feedback
link in the ui - Visit the main repository for documentation
- File issues in the project's issue tracker