Phase 1 / Ep 02: macOS Native Installation —— 5-Minute Quick Start

⏱ Est. reading time: 5 min Updated on 4/13/2026

🎯 Learning Objective: Complete the OpenClaw installation and initial onboarding on macOS, and verify that the Gateway is running normally.

1. Prerequisite Checks

Before starting, ensure your Mac meets the following requirements:

Dependency Minimum Version Check Command
macOS 12 Monterey+ sw_vers
Node.js v18.0+ node -v
Git 2.30+ git --version
curl Any curl --version

If Node.js is missing, it is recommended to install it using nvm:

curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.0/install.sh | bash
nvm install 22

2. One-Click Installation

Open Terminal and execute:

curl -fsSL https://openclaw.ai/install.sh | bash

The installation script will:

  1. Download the OpenClaw CLI to ~/.openclaw/bin/
  2. Add the openclaw command to your PATH
  3. Automatically launch the Onboarding wizard

3. Complete Onboarding Wizard Process

After installation, the wizard will guide you through four core steps:

graph LR
    A["1️⃣ Select LLM\nAnthropic / OpenAI\n/ Ollama"] --> B["2️⃣ Enter API Key\nor configure local model"]
    B --> C["3️⃣ Select Channel\nTelegram / Discord\n/ Configure later"]
    C --> D["4️⃣ Daemon Config\nAuto-start on boot · Background run"]

Step 1: Select LLM Provider

? Select your AI provider:
  ❯ Anthropic (Claude)
    OpenAI (GPT)
    Ollama (Local)
    Other

Recommended: Select Anthropic, as Claude performs best in tool calling.

Step 2: Enter API Key

? Enter your Anthropic API Key: sk-ant-api03-xxxx
✅ API Key validated successfully!

⚠️ The API Key is only stored locally in the ~/.openclaw/ directory and is never uploaded to any external servers.

Step 3: Select Messaging Channel

For the initial installation, you can choose "Configure later" (Skip). We will explain Telegram integration in detail in Ep 06.

Step 4: Daemon Configuration

? Install as background daemon (auto-start on login)?
  ❯ Yes (recommended)
    No (manual start)

By selecting Yes, OpenClaw will create a macOS LaunchAgent to enable auto-start on boot.

4. Initial Health Check

Once the installation is complete, run two verification commands:

# Check Gateway status
openclaw status

# Automatically diagnose and fix issues
openclaw doctor

You should see output similar to:

🦀 OpenClaw Status
══════════════════════════════
Gateway:    ● Running (PID 12345)
Agent:      personal (default)
LLM:        Anthropic / claude-sonnet-4-20250514
Channels:   None configured
Uptime:     2m 34s

5. Key Commands Quick Reference

Step Command Description
Install curl -fsSL https://openclaw.ai/install.sh | bash One-click installation
Wizard openclaw onboard --install-daemon Manually restart Onboarding
Status openclaw status Health check
Diagnose openclaw doctor Auto-fix issues
Update openclaw update Upgrade to the latest version

Next Episode Preview: In Ep 03, we will containerize OpenClaw using Docker to achieve more secure process-level isolation. If you plan to run the Agent long-term on a VPS, this method is highly recommended.