Skip to main content
Validates your .env secrets, configures the Pulumi stack, and provisions infrastructure for your agent team — all in a single command.

​
Usage

​
Options

​
What It Does

  1. Validates secrets — Loads .env, resolves ${env:VAR} references, validates formats, and reports any missing secrets
  2. Fetches identities — Resolves agent identities from Git repos or local paths
  3. Auto-resolves secrets — Fetches Linear user UUIDs and other auto-resolvable values
  4. Configures Pulumi — Creates/selects the stack and sets all config values
  5. Provisions infrastructure — Runs pulumi up to create servers, volumes, networking, and security groups
  6. Bootstraps agents — Installs OpenClaw, tools, and configured integrations on each agent
  7. Connects to Tailscale — Joins agents to your Tailscale network
In previous versions, steps 1-4 were handled by a separate clawup setup command. As of v2.6.0, deploy handles everything automatically.

​
Local Docker Mode

Use --local to run agents in local Docker containers for testing:
This creates a separate Pulumi stack (<stackName>-local), copies config from your cloud stack (skipping Tailscale keys), and runs agents locally. No cloud resources are created.

​
Examples

First deployment takes 3-5 minutes. Subsequent deployments are incremental — only changes are applied.
Deployment creates cloud resources that incur costs. Review your config and cost estimates before deploying.