Skip to main content
A deployment in ClawControl represents a complete OpenClaw instance running on a cloud VPS. Each deployment includes the server infrastructure, OpenClaw installation, and all configuration needed to run your AI agent.

​
What is a Deployment?

A deployment is a managed OpenClaw instance that ClawControl creates and maintains on your chosen cloud provider. When you create a deployment, ClawControl:
  1. Provisions a VPS server on your chosen provider
  2. Configures the operating system and dependencies
  3. Installs and configures OpenClaw
  4. Sets up networking and security (SSH, Tailscale)
  5. Starts the OpenClaw daemon
Each deployment is completely isolated with its own server, SSH keys, and configuration. You can run multiple deployments simultaneously.

​
Deployment Configuration

Every deployment has a configuration file (config.json) that defines its settings: 
interface DeploymentConfig {
  name: string;                    // Unique deployment name
  provider: "hetzner" | "digitalocean" | "vultr";
  createdAt: string;
  hetzner?: HetznerConfig;         // Provider-specific config
  digitalocean?: DigitalOceanConfig;
  openclawConfig?: OpenClawConfig; // Custom OpenClaw settings
  openclawAgent?: OpenClawAgentConfig; // AI provider & channel
  skipTailscale?: boolean;         // Optional: skip Tailscale setup
}

​
Example Configuration

{
  "name": "my-agent",
  "provider": "hetzner",
  "createdAt": "2025-01-15T10:30:00.000Z",
  "hetzner": {
    "apiKey": "***",
    "serverType": "cpx11",
    "location": "ash",
    "image": "ubuntu-24.04"
  },
  "openclawAgent": {
    "aiProvider": "openrouter",
    "aiApiKey": "***",
    "model": "moonshotai/kimi-k2.5",
    "channel": "telegram",
    "telegramBotToken": "***"
  }
}

​
Deployment State

ClawControl tracks each deployment’s state in a separate state.json file: 
interface DeploymentState {
  status: DeploymentStatus;
  serverId?: string;              // Provider's server ID
  serverIp?: string;              // Public IP address
  tailscaleIp?: string;           // Tailscale VPN IP
  sshKeyId?: string;              // Provider's SSH key ID
  sshKeyFingerprint?: string;
  checkpoints: Checkpoint[];      // Recovery checkpoints
  gatewayToken?: string;          // Dashboard auth token
  lastError?: string;
  deployedAt?: string;
  updatedAt: string;
}

​
Deployment Status

Deployments progress through these statuses:
  • initialized - Configuration created, ready to deploy
  • provisioning - Creating server resources
  • configuring - Installing and configuring software
  • deployed - Successfully running
  • failed - Deployment encountered an error
  • updating - Applying changes to existing deployment
{
  "status": "deployed",
  "serverId": "12345678",
  "serverIp": "203.0.113.42",
  "tailscaleIp": "100.64.1.5",
  "sshKeyFingerprint": "SHA256:abc123...",
  "checkpoints": [
    {
      "name": "server_created",
      "completedAt": "2025-01-15T10:32:00.000Z",
      "retryCount": 0
    },
    {
      "name": "ssh_connected",
      "completedAt": "2025-01-15T10:34:00.000Z",
      "retryCount": 0
    },
    // ... more checkpoints
    {
      "name": "completed",
      "completedAt": "2025-01-15T10:45:00.000Z",
      "retryCount": 0
    }
  ],
  "deployedAt": "2025-01-15T10:45:00.000Z",
  "updatedAt": "2025-01-15T10:45:00.000Z"
}

​
Deployment Lifecycle

A typical deployment goes through these phases:

​
1. Creation

Define your deployment configuration using a template or custom settings: 
clawcontrol new
This creates a deployment directory at ~/.clawcontrol/deployments/<name>/ containing:
  • config.json - Your deployment configuration
  • state.json - Current deployment state
  • ssh/ - SSH key pair for server access

​
2. Deployment

Execute the deployment to provision and configure your server: 
clawcontrol deploy <name>
ClawControl uses a checkpoint-based system to ensure reliable deployment with automatic recovery.

​
3. Operation

Once deployed, you can:
  • View status: clawcontrol status <name>
  • Access dashboard: clawcontrol dashboard <name>
  • SSH into server: clawcontrol ssh <name>
  • View logs: clawcontrol logs <name>

​
4. Destruction

When you’re done, remove the deployment: 
clawcontrol destroy <name>
This deletes the VPS server, removes cloud resources, and cleans up local configuration.

​
Storage Location

All deployments are stored in ~/.clawcontrol/deployments/: 
~/.clawcontrol/
└── deployments/
    ├── my-agent/
    │   ├── config.json    # Deployment configuration
    │   ├── state.json     # Current state & checkpoints
    │   └── ssh/
    │       ├── id_ed25519     # Private SSH key
    │       └── id_ed25519.pub # Public SSH key
    └── another-agent/
        ├── config.json
        ├── state.json
        └── ssh/
SSH keys in the deployment directory provide root access to your servers. Keep the ~/.clawcontrol/ directory secure and never commit it to version control.

​
Full Deployment Interface

From the source code at src/types/index.ts:204-208: 
interface Deployment {
  config: DeploymentConfig;
  state: DeploymentState;
  sshKeyPath: string;
}
This combined interface represents everything ClawControl knows about a deployment, merging configuration with runtime state.

​
Best Practices

Use descriptive names

Choose deployment names that indicate their purpose, like production-agent or dev-testing.

One deployment per purpose

Keep production, development, and testing agents in separate deployments.

Monitor deployment status

Regularly check clawcontrol status to ensure your agents are running.

Backup configurations

Back up your ~/.clawcontrol/deployments/ directory to preserve settings and SSH keys.

​
Next Steps

Templates

Learn how to use templates for faster deployment

Checkpoints

Understand the checkpoint-based recovery system

Deploy Command

See the complete deploy command reference

Status Command

Check your deployment status

Build docs developers (and LLMs) love

Get started for freeTalk to us