Production-grade AI orchestration platform with unified installation
Automagik Suite is a comprehensive AI workflow automation platform that combines multiple services into a cohesive ecosystem. This repository provides a streamlined installation process that sets up the entire suite with minimal configuration.
The Automagik Suite uses a hybrid architecture:
- Infrastructure: Docker containers with host networking for databases and message queues
- Application Services: PM2-managed processes for better performance and native integration
- Optional Services: LangFlow and Evolution API can be enabled during installation
- automagik - Main AI orchestrator and agent management (Port: 8881)
- automagik-spark - Workflow engine with Celery workers (Port: 8883)
- automagik-tools - MCP tools hub with SSE/HTTP endpoints (Ports: 8884/8885)
- automagik-omni - Multi-tenant hub for instance management (Port: 8882)
- automagik-ui - Next.js web interface (Port: 8888)
- langflow - Visual AI workflow builder (Port: 7860)
- automagik-evolution - WhatsApp API integration (Port: 8080)
- PostgreSQL databases for each service (Ports: 5401-5403)
- Redis instances for caching and queues (Ports: 5411-5413)
- RabbitMQ for Evolution API messaging (Port: 5431)
- Ubuntu/Debian or macOS
- Git
- Sudo access (for system dependencies only)
Choose your preferred installation method:
source <(curl -sSL https://raw.githubusercontent.com/namastexlabs/automagik-start/main/install-and-cd.sh)
This method automatically changes to the automagik directory after installation.
curl -sSL https://raw.githubusercontent.com/namastexlabs/automagik-start/main/start.sh | bash
cd automagik # Manual directory change required
git clone https://github.com/namastexlabs/automagik-start.git automagik
cd automagik
./install.sh
The installer will:
- Install system dependencies (Python 3.12, Node.js 22 LTS, Docker, PM2)
- Prompt for optional services (LangFlow, Evolution API)
- Clone all service repositories
- Set up Docker infrastructure
- Build and configure all services
- Start everything with PM2
All configuration is managed through a single .env
file in the root directory:
# Copy the example file
cp .env.example .env
# Edit with your API keys
nano .env
Key configurations:
- AI Providers: OpenAI, Anthropic, Google Gemini API keys
- Database URLs: All use localhost with specific ports
- Service URLs: All services accessible on localhost
- API Keys: Internal service authentication keys
Services are managed by PM2 using ecosystem.config.js
:
# View all services
make status
# Start all services
make start
# Stop all services
make stop
# View logs
make logs
# Restart specific service
make restart-agents
make install
- Complete installation (infrastructure + services)make start
- Start everythingmake stop
- Stop everythingmake restart
- Restart all servicesmake update
- Git pull and restartmake logs
- View colorized logsmake status
- Check service status
make start-[service]
- Start specific service (agents, spark, tools, omni, ui)make stop-[service]
- Stop specific servicemake restart-[service]
- Restart specific servicemake logs-[service]
- View specific service logs
make start-langflow
- Start LangFlowmake stop-langflow
- Stop LangFlowmake start-evolution
- Start Evolution APImake stop-evolution
- Stop Evolution API
make start-infrastructure
- Start Docker containersmake stop-infrastructure
- Stop Docker containers
All Docker services use host network mode for simplified networking:
- No port mapping required
- Direct access to localhost
- Full LAN connectivity (e.g., 192.168.x.x)
- Better performance
Docker compose files:
docker-infrastructure.yml
- Core databases and Redisdocker-langflow.yml
- LangFlow service (optional)docker-evolution.yml
- Evolution API and dependencies (optional)
automagik-suite/
βββ .env # Main configuration file
βββ .env.example # Configuration template
βββ Makefile # Service orchestration commands
βββ ecosystem.config.js # PM2 configuration
βββ install.sh # Installation script
βββ docker-infrastructure.yml # Core Docker services
βββ docker-langflow.yml # LangFlow Docker service
βββ docker-evolution.yml # Evolution API Docker services
βββ scripts/ # Utility scripts
β βββ utils/ # Shared utilities
β βββ system/ # OS-specific installers
β βββ deploy/ # Deployment scripts
βββ automagik/ # Main orchestrator
βββ automagik-spark/ # Workflow engine
βββ automagik-tools/ # MCP tools
βββ automagik-omni/ # Multi-tenant hub
βββ automagik-ui/ # Web interface
- Main AI orchestration service
- Agent management and routing
- PostgreSQL database on port 5401
- Redis cache on port 5411
- Workflow automation engine
- Celery task processing
- PostgreSQL database on port 5402
- Redis queue on port 5412
- MCP (Model Context Protocol) tools
- SSE endpoint on port 8884
- HTTP endpoint on port 8885
- No database required
- Multi-tenant instance management
- API gateway for multiple services
- Uses spark database
- Next.js 15 web interface
- Real-time updates via SSE
- Production build served by PM2
# Start in development mode
make dev
# Follow logs
make logs-follow
# Check specific service
make status-spark
# Pull latest changes
make pull
# Update and restart
make update
- Clone repository to root directory
- Add to
ecosystem.config.js
- Update Makefile targets
- Run
make install-[service]
Port Conflicts
# Check what's using a port
lsof -i :8881
# Kill process using port
kill -9 $(lsof -t -i:8881)
PM2 Issues
# Reset PM2
pm2 kill
pm2 resurrect
# Clear logs
pm2 flush
Docker Issues
# Reset Docker containers
make stop-infrastructure
docker system prune -a
make start-infrastructure
- PM2 logs:
~/.pm2/logs/
- Service logs:
[service]/logs/
- Install log:
automagik-install.log
- Infrastructure:
make status-infrastructure
- Services:
make status
- Specific service:
curl http://localhost:[port]/health
- Main Orchestrator: http://localhost:8881
- Workflow Engine: http://localhost:8883
- Tools SSE: http://localhost:8884
- Tools HTTP: http://localhost:8885
- Multi-tenant Hub: http://localhost:8882
- Web Interface: http://localhost:8888
- LangFlow (optional): http://localhost:7860
- Evolution API (optional): http://localhost:8080
All services use API key authentication via x-api-key
header.
Default key: namastex888
(configure in .env
)
- Change default API keys in production
- Use strong database passwords
- Configure firewall rules for exposed ports
- Enable HTTPS in production environments
- Regularly update dependencies
MIT License - see LICENSE file for details.
Built with β€οΈ by NamasteX Labs