How It Works

From sign-up to a running, secured AI agent in under 5 minutes. Here's exactly what happens.

From sign-up to running agent

1

Choose your plan and configure your agent

Select a tier that fits your needs, name your agent, and choose which messaging channels to enable. If you're on a Pro or Enterprise plan, pick an email identity for your agent.

2

Dedicated server provisioned automatically

Within seconds of payment, our provisioning pipeline allocates a dedicated virtual private server exclusively for your agent. No shared infrastructure — your server is yours alone.

3

Security hardening applied

All 7 security layers are applied automatically: firewall rules, encrypted storage volumes, container sandboxing, localhost-only gateway binding, token authentication, DM pairing policies, and tool allowlists.

4

Encrypted tunnel established

Your agent establishes an outbound-only encrypted tunnel to our edge network. No listening ports are opened — your agent is invisible to the public internet. Only our control plane can communicate with it.

5

Agent goes live

Your AI agent comes online and is ready to accept messages. Connect your messaging channels (Telegram, Discord, Slack, WhatsApp) and your agent begins responding to approved contacts.

Your agent's dedicated environment

Every ClawTrust agent runs in complete isolation. Nothing is shared with other customers.

Dedicated virtual private server

Every agent gets its own VPS with dedicated CPU, memory, and storage. No multi-tenancy at the infrastructure level.

Encrypted storage

Data volumes are encrypted at rest using AES-256. Even physical access to the server hardware cannot expose your data.

Containerized execution

Your agent and all tool executions run inside isolated containers with strict resource limits and read-only filesystem access.

Network isolation

Zero public-facing ports. All communication happens through outbound-only encrypted tunnels initiated by your agent.

How your agent communicates

All messages flow through encrypted channels. Your agent never exposes a public endpoint.

Messaging Platform
Edge Network
Encrypted Tunnel
Your Agent

Your agent initiates the tunnel connection outbound. No inbound ports are ever opened.

What happens behind the scenes

Once your agent is running, we handle the operational work so you don't have to.

Continuous health monitoring

We monitor your agent's health around the clock. If your agent becomes unresponsive, automated recovery kicks in within seconds.

Automatic security patching

When security updates are released, they're deployed fleet-wide through automated rolling updates — no downtime, no manual intervention.

Graceful recovery

If a process crashes, the agent automatically restarts. Persistent state is preserved through encrypted storage, so no conversations are lost.

Ready to hire your first AI employee?

Production-ready in 5 minutes. All 7 security layers included on every plan.