Cyberpunk Production Deployment System - Self-healing Docker orchestration with 7 containers, AI-powered multi-agent routing, vector memory, and a neon-green glassmorphism UI.
git clone https://github.com/Mr-Process/t-rex-claw.git
cd t-rex-claw
sudo bash install.shThat is it. The installer handles Docker, containers, AI models, SSL, and the self-healing daemon.
Internet
|
[ Traefik ] :80 / :443 / :8080
/ | \
Landing Page API Control N8N
:3000 :8000 :8081 :5678
|
[ Syve Core ]
/ \
[ Ollama ] [ Qdrant ]
:11434 :6333
| Service | Container | Port | Description |
|---|---|---|---|
| Traefik | trex-traefik | 80, 443, 8080 | Reverse proxy, SSL, rate limiting |
| OpenClaw-Syve | trex-syve | 8000 | Core AI engine (FastAPI) |
| OpenClaw-Landing | trex-landing | 3000 | Cyberpunk landing page |
| OpenClaw-Control | trex-control | 8081 | Control center dashboard |
| Ollama | trex-ollama | 11434 | Local LLM engine |
| Qdrant | trex-qdrant | 6333 | Vector database |
| N8N | trex-n8n | 5678 | Workflow automation |
The scripts/autolife.sh daemon runs every minute via cron, monitoring all 7 containers. If any container stops or becomes unhealthy, it auto-restarts it, cleans up disk when usage exceeds 90%, and rotates logs.
Four specialized agents powered by local Ollama models:
| Agent | Model | Purpose |
|---|---|---|
| Messaging | qwen2.5:7b | Natural language conversations |
| Code | deepseek-coder-v2 | Code generation and analysis |
| Research | llama3:8b | Research and knowledge tasks |
| Task | mistral-nemo | Task planning and execution |
- LanceDB - Local embedded vector store
- Qdrant - High-performance similarity search
- Automatic SSL via Let's Encrypt
- HTTP to HTTPS redirect
- Security headers (XSS, HSTS, CSP)
- Rate limiting (100 req/min)
- Gzip compression
Glassmorphism design with neon green (#00ff88) theming, matrix rain animations, and real-time system monitoring dashboards.
- OS: Ubuntu 20.04+ / Debian 11+ / CentOS 8+ (any Linux with systemd)
- RAM: 16 GB minimum (32 GB recommended for all LLM models)
- Disk: 50 GB free space
- GPU: NVIDIA GPU recommended for Ollama (optional, CPU works too)
Docker and Docker Compose are installed automatically by the installer.
git clone https://github.com/Mr-Process/t-rex-claw.git && cd t-rex-claw && sudo bash install.sh-
Clone the repo
git clone https://github.com/Mr-Process/t-rex-claw.git cd t-rex-claw -
Configure environment (optional, defaults work for localhost)
cp .env.template .env nano .env
-
Run the installer
sudo bash install.sh
The installer will:
- Install Docker and Docker Compose (if not present)
- Create
.envfrom template (if not exists) - Build and start all 7 containers
- Pull Ollama AI models
- Set up the Autolife cron job
All settings are in .env (copy from .env.template):
| Variable | Default | Description |
|---|---|---|
DOMAIN |
localhost | Your domain for SSL/routing |
ACME_EMAIL |
admin@example.com | Email for Let's Encrypt |
OLLAMA_MODEL |
qwen2.5:7b | Default Ollama model |
SYVE_PORT |
8000 | API engine port |
LANDING_PORT |
3000 | Landing page port |
CONTROL_PORT |
8081 | Control center port |
N8N_PORT |
5678 | N8N workflow port |
QDRANT_PORT |
6333 | Vector database port |
N8N_BASIC_AUTH_USER |
admin | N8N login username |
N8N_BASIC_AUTH_PASSWORD |
changeme | N8N login password |
AUTOLIFE_INTERVAL |
30 | Health check interval (sec) |
BACKUP_RETENTION_DAYS |
7 | Days to keep backups |
sudo bash install.sh status # Show container status
sudo bash install.sh restart # Restart all services
sudo bash install.sh stop # Stop all services
sudo bash install.sh logs # Follow container logs
sudo bash install.sh uninstall # Remove everythingsudo bash scripts/health-check.sh # Quick health verification
sudo bash scripts/backup.sh # Manual backup
sudo bash scripts/autolife.sh # Run self-healing checkt-rex-claw/
|-- install.sh # Master installer
|-- docker-compose.production.yml # All 7 services
|-- .env.template # Environment config template
|-- .gitignore
|-- LICENSE
|-- README.md
|-- CONTRIBUTING.md
|-- CHANGELOG.md
|-- traefik/
| |-- traefik.yml # Traefik static config
| +-- dynamic/
| +-- security.yml # Security middleware
|-- services/
| |-- syve/ # Core AI engine
| | |-- Dockerfile
| | |-- main.py # FastAPI application
| | |-- requirements.txt
| | |-- core/
| | | |-- __init__.py
| | | |-- agents.py # Multi-agent router
| | | +-- memory.py # Vector memory system
| | +-- public/
| | +-- index.html # Dashboard UI
| |-- landing/
| | |-- nginx.conf
| | +-- html/
| | +-- index.html # Landing page
| +-- control/
| |-- nginx.conf
| +-- html/
| +-- index.html # Control center
|-- scripts/
| |-- autolife.sh # Self-healing daemon
| |-- health-check.sh # Health verification
| +-- backup.sh # Backup automation
+-- data/ # Runtime data (gitignored)
+-- acme/ # SSL certificates
| Method | Endpoint | Description |
|---|---|---|
| GET | /api/v1/health |
Health check |
| GET | /api/v1/system/stats |
CPU, memory, disk, uptime |
| GET | /api/v1/agents |
List AI agents |
| GET | /api/v1/memory/stats |
Vector memory status |
| POST | /api/v1/chat |
Chat with AI agents |
Containers not starting?
sudo bash install.sh logs
docker ps -aPort conflicts?
Edit .env to change port assignments, then sudo bash install.sh restart.
Ollama models not loading?
docker exec trex-ollama ollama list
docker exec trex-ollama ollama pull qwen2.5:7bSSL not working? Ensure your domain DNS points to the server and ports 80/443 are open.
See CONTRIBUTING.md for guidelines.
MIT License. See LICENSE for details.
Built by Mr-Process