This is the clean baseline install flow. You want the CLI installed, the gateway configured in local mode, the dashboard reachable, and the model layer verified before you expand the setup.
Quick answer
Install the CLI, run onboarding, verify the gateway, open the dashboard, and confirm the model status before you customize anything else.
Command line steps
1. Install the CLI
Use the standard installer so the rest of the docs and commands line up.
BASH
$curl -fsSL https://openclaw.ai/install.sh | bash
2. Run onboarding
Let OpenClaw create the config and initial gateway state.
BASH
$openclaw onboard
3. Verify the basics
If these checks pass, you have a clean starting point.
BASH
$openclaw gateway status --require-rpc
openclaw dashboard
openclaw models status
What to check if it still fails
- If
gateway statusfails, debug that first. Do not add channels on top of a broken service layer. - If the dashboard opens but replies fail, inspect model auth with
openclaw models status --check. - If onboarding feels wrong because you copied older config snippets, run
openclaw doctor --fixand repeat the setup flow cleanly.