Security built into the architecture
Your OpenClaw runs on an isolated VM behind our managed network. No public IP, no direct internet exposure. Security isn't an afterthought -- it's the foundation.
Your instance never touches the public internet
Traffic flows through our managed reverse proxy. Your VM has a private IP only.
No public IP
Your VM only has a private IP. Not scannable, not directly reachable.
Managed proxy
All traffic routes through our managed reverse proxy with automatic SSL.
Managed updates
We handle security patches and infrastructure updates.
This is why architecture matters
OpenClaw should never be on the public internet
You can configure a VPS to be secure. But it requires constant management, and you can't afford a single mistake.
The problem with public VPSes
Even when configured correctly
Your instance is directly attached to the internet
A public IP means your server is indexed by automated scanners within hours. You're responsible for every firewall rule, every port, every update.
One missed update, one misconfiguration
Security isn't a one-time setup. It requires constant vigilance: patching, monitoring, hardening. People forget. People make mistakes. That's human.
The right answer is to remove the attack surface entirely
Instead of hardening a public-facing server and hoping nothing slips through, put the instance behind a private network where it can't be reached directly.
Tiller instances have no public IP. There's nothing to scan, nothing to find. That's the architecture OpenClaw deserves.
How Tiller protects you
You can't stop prompt injection at the AI level. But you can limit what happens when it succeeds.
Network isolation
Your VM has a private IP only. Not scannable, not directly reachable from the internet. All traffic routes through our managed proxy.
Gateway locked down
Your OpenClaw gateway only accepts local connections. External access requires authentication through our secure proxy on the private network.
Brute force protection
Fail2ban monitors for suspicious activity and auto-bans attacking IPs before they can do damage.
Firewall by default
UFW blocks everything except SSH and HTTPS. Even misconfigurations can't expose your instance to the internet.
HTTPS everywhere
Automatic SSL on every instance. All traffic is encrypted in transit. No manual certificate management.
SSH hardened
Password auth disabled. Root login blocked. Key-only access. The basics most DIY setups skip.
Ready to deploy on secure, isolated infrastructure?