Installation & Deployment

📋 Prerequisites & Recommended Configuration

Recommended Configuration

For the best experience, we recommend:

Operating System:

• 🍎 Mac (Highly Recommended): Best native support, can operate Calendar, Notes, screenshots, and other system features
• 🪟 Windows: Fully usable, but some system integration features are limited
• 🐧 Linux: Great for developers, flexible configuration

IM Tool Selection:

• 🌍 International Users: Recommended Telegram (best compatibility, most complete features)
• 🇨🇳 China Users: Recommended Lark (modern, developer-friendly, feature-rich)
• Alternatives: WeCom, DingTalk, QQ

Deployment Method:

• 💻 Have a Mac: Recommended local deployment (best experience, full features)
• ☁️ No Mac or want 24/7 operation: Recommended cloud deployment (low cost, stable and reliable)

Why Mac is Recommended?

OpenClaw works best on Mac because:

• ✅ Best native support with deep system integration
• ✅ Can operate Mac Calendar, Notes, Reminders
• ✅ Perfect screenshot support
• ✅ Seamless sync with iPhone and iPad
• ✅ Smarter file management
• ✅ Simple development environment setup

Why Lark is Recommended (China)?

• ✅ Modern design, great user experience
• ✅ Developer-friendly with comprehensive API
• ✅ Supports rich text, documents, spreadsheets
• ✅ Stable message push notifications
• ✅ Feature-rich free tier

Why Telegram is Recommended (International)?

• ✅ Large global user base
• ✅ Most comprehensive API, most powerful features
• ✅ Rich Bot functionality support
• ✅ Real-time message push
• ✅ Good privacy protection

Quick Navigation

Recommended Paths:

• 🍎 Have a Mac → Mac Local Deployment + Lark Configuration
• ☁️ No Mac / Want 24/7 operation → Cloud One-Click Deployment + Lark Configuration

All Deployment Methods:

• 🍎 Mac Local Deployment (Recommended)
• 🪟 Windows Local Deployment
• 🐧 Linux Local Deployment
• 🚀 Cloud One-Click Deployment
• 🇨🇳 China One-Click Installation
• ☁️ Cloudflare Workers Deployment (Advanced)
• 🐳 Docker Deployment (Optional)

Configuration Guides:

• 🔑 API Configuration Guide
• 🔄 Version Upgrade Guide
• ❓ Common Issues & Solutions

Mac Local Deployment (Recommended)

Why Choose Mac Local Deployment?

Advantages:

• ✅ System Integration: Can operate Calendar, Notes, file system
• ✅ Privacy & Security: Data stays completely local, not uploaded to the cloud
• ✅ Fast Response: Runs locally, no network latency
• ✅ Full Features: Supports all advanced features
• ✅ Low Cost: No need to purchase cloud servers
• ✅ Developer-Friendly: Easy to debug and customize

Ideal For:

• Users with a Mac
• Privacy-conscious users
• Users who need system integration features
• Developers and tech enthusiasts

System Requirements

Hardware Requirements:

• CPU: M-series chip or Intel i5 or above
• RAM: 8GB or more (16GB recommended)
• Disk: 10GB or more free space

System Version:

• macOS 12 Monterey or later
• Recommended macOS 14 Sonoma or macOS 15 Sequoia

Prerequisites:

• Node.js 22.0.0+ (will be installed automatically)
• Homebrew (optional, for installing dependencies)

Installation Steps

Step 1: Open Terminal

1. Press Command + Space to open Spotlight
2. Type Terminal
3. Press Enter to open Terminal

Step 2: Install OpenClaw

Run the following command in Terminal:

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

The installation process will automatically:

• Detect system environment
• Install Node.js (if not installed)
• Download OpenClaw
• Configure environment variables

Estimated Time: 2-5 minutes

Step 3: Verify Installation

After installation, run the following command to verify:

openclaw --version

If a version number is displayed (e.g., 2026.3.2), the installation was successful!

Step 4: Initialize Configuration

Run the configuration wizard:

openclaw onboard

Configuration Flow:

1. Accept Risk Notice:

Select Yes to continue

2. Select Startup Mode:

Recommended to select QuickStart:

3. Select AI Model:

Choose your AI provider (supports major domestic and international models):

Recommended for China users:

• Kimi (Moonshot AI): Long text expert, 2 million character context
• DeepSeek: Best value for money, strong reasoning capabilities
• GLM (Zhipu AI): Good Chinese comprehension, multimodal support

4. Enter API Key:

Enter the corresponding API Key based on your selected model (see API Configuration Guide)

5. Select Chat Tool:

• If connecting to Lark/Telegram, select the corresponding option
• If not connecting yet, select None (can be configured later)

6. Gateway Port Setting:

Default 18789 is fine:

7. Select Skills:

Use the spacebar to select the skills you need, or skip directly:

8. API Key Configuration:

If you don't have one, select no to skip:

9. Enable Hooks:

Recommended to enable these three hooks (for content guidance, logging, and session recording):

10. Complete Configuration:

After configuration is complete, the Gateway service will automatically start and open the Web UI (http://127.0.0.1:18789/chat)

Step 5: Verify Installation

# Check Gateway status
openclaw channels status

# Should display:
# Gateway reachable.

Daily Usage

Start OpenClaw:

# Start Gateway service
openclaw gateway start

# Or use systemd (recommended, auto-start on boot)
openclaw gateway enable

Access Web UI:

Open your browser and visit: http://127.0.0.1:18789/chat

Stop Service:

openclaw gateway stop

Connect to Lark (Recommended)

After Mac local deployment, we strongly recommend connecting to Lark for the best experience:

1. Refer to Chapter 9: Lark Bot Configuration
2. Once configured, you can chat with OpenClaw anytime in Lark
3. Supports text, images, files, and other message types

Common Issues

Q1: Permission denied during installation?

# Install with sudo
curl -fsSL https://openclaw.ai/install.sh | sudo bash

Q2: How to update OpenClaw?

openclaw update

Q3: How to uninstall?

openclaw uninstall

Windows Local Deployment

System Requirements

Hardware Requirements:

• CPU: 2 cores or more
• RAM: 4GB or more (8GB recommended)
• Disk: 10GB or more free space

Operating System:

• Windows 10 or Windows 11

Prerequisites:

• Node.js 22.0.0+

Deployment Method Selection

There are two deployment methods for Windows:

1. WSL2 + Ubuntu (Highly Recommended): Officially recommended method, provides full Linux environment support
2. PowerShell Native Deployment: Pure Windows environment, suitable for users who don't want to use WSL2

Method 1: WSL2 + Ubuntu Deployment (Highly Recommended)

This is the officially recommended Windows deployment method, providing the most complete Linux environment support.

Step 1: Enable WSL2

Open PowerShell as Administrator and run:

# Enable WSL feature
dism.exe /online /enable-feature /featurename:Microsoft-Windows-Subsystem-Linux /all /norestart
dism.exe /online /enable-feature /featurename:VirtualMachinePlatform /all /norestart

# Set WSL 2 as default version
wsl --set-default-version 2

Restart your computer.

Step 2: Install Ubuntu

Method 1: Install from Microsoft Store (Recommended)

1. Open Microsoft Store
2. Search for "Ubuntu 22.04 LTS" or "Ubuntu 24.04 LTS"
3. Click "Get" and install
4. Set username and password on first launch

After installation, the Ubuntu terminal will open automatically. Follow the prompts to set your username and password.

Step 3: Update Ubuntu System

Run in the Ubuntu terminal:

# Update package list
sudo apt update && sudo apt upgrade -y

# Install basic tools
sudo apt install -y curl git wget build-essential

Step 4: Install Node.js 22+

# Add NodeSource repository
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -

# Install Node.js
sudo apt install -y nodejs

# Verify version (must be ≥22.x)
node -v
npm -v

Step 5: Install OpenClaw

Method A: One-click script installation

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

Step 6: Verify Installation

# Check version
openclaw --version

# View help
openclaw --help

# Check system status
openclaw status

Step 7: Configure Windows Access to WSL2 Service

Since OpenClaw runs in WSL2, you need to configure port forwarding for Windows access.

Create startup script start-openclaw.bat:

@echo off
echo Starting OpenClaw Gateway in WSL2...
wsl -d Ubuntu-22.04 -u root service openclaw start
timeout /t 3
start http://localhost:18789

Or start directly in WSL2:

# In WSL2 Ubuntu terminal
openclaw gateway run --port 18789

Then visit http://localhost:18789 in your Windows browser

Method 2: PowerShell Native Deployment

Suitable for pure Windows users who don't want to use WSL2.

Step 1: Install Node.js 22+

Method 1: Download from Official Website

1. Visit https://nodejs.org/zh-cn
2. Download the Windows installer (LTS version 22.x)
3. Run the installer, check "Automatically install the necessary tools"

Step 2: Verify Node.js Installation

# Open PowerShell
node -v
npm -v

Step 3: Install OpenClaw as Administrator

Important: You must run PowerShell as Administrator.

# Install latest stable version
npm install -g openclaw@latest

# Or install the Chinese localized version
npm install -g @qingchencloud/openclaw-zh@latest

Step 4: Resolve Installation Permission Issues

If you encounter permission errors:

# Method A: Enable PowerShell script execution
Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser

# Method B: Change npm installation directory
npm config set prefix "C:\npm"
npm config set cache "C:\npm-cache"

# Add directory to PATH
[Environment]::SetEnvironmentVariable("Path", $env:Path + ";C:\npm", "User")

Step 5: Verify Installation

openclaw --version
openclaw --help

Step 6: Resolve Common Issues

Issue: sharp module loading failure

# Clear npm cache
npm cache clean --force

# Reinstall
npm install -g openclaw@latest --force

Issue: Windows Defender blocking

Add the OpenClaw installation directory to Windows Defender exclusions:

C:\Users\YourUsername\AppData\Roaming\npm
C:\Users\YourUsername\.openclaw

Initialize Configuration

After installation, you need to run the initialization wizard.

Start the Initialization Wizard

openclaw onboard --install-daemon

Configure AI Model Provider

OpenClaw needs to connect to an AI model to work.

Example: Configuring Anthropic Claude (Recommended):

# WSL2 or PowerShell
openclaw models auth add
# Follow prompts to select anthropic
# Enter API Key: sk-ant-xxx

Bind Messaging Channels

1. Telegram

Create a Bot:

1. Search for @BotFather in Telegram
2. Send /newbot to create a bot
3. Save the Bot Token

Configure:

openclaw channels add telegram
openclaw config set channels.telegram.botToken "your-bot-token"
openclaw gateway restart

2. WhatsApp

# Login to WhatsApp (displays QR code)
openclaw channels login whatsapp

# Scan with your phone's WhatsApp

3. WeCom (Recommended for China)

# Install WeCom plugin
openclaw plugins install @m1heng-clawd/wework

# Configure
openclaw config set channels.wework '{"enabled":true,"corpId":"xxx","agentSecret":"xxx"}' --json

4. Lark (Recommended for China)

# Install Lark plugin
openclaw plugins install @m1heng-clawd/feishu

# Configure
openclaw config set channels.feishu '{"enabled":true,"appId":"cli_xxx","appSecret":"xxx"}' --json

Windows Command Quick Reference

System Management:

Configuration Management:

Linux Local Deployment

System Requirements

Recommended Distributions:

• Ubuntu 20.04+
• Debian 11+
• CentOS 8+

Installation Steps

Step 1: Install Node.js

# Ubuntu/Debian
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -
sudo apt-get install -y nodejs

# Verify installation
node --version

Step 2: Install OpenClaw

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

Step 3: Verify Installation

openclaw --version

Step 4: Initialize Configuration

openclaw onboard

2.1 System Requirements & Preparation

Cloud Deployment Requirements

If you choose cloud deployment, no local environment is needed, you only need:

• ✅ A web browser
• ✅ ~20 CNY/month budget
• ✅ 10 minutes of time

Cloud One-Click Deployment

Why Choose Cloud Deployment?

Cloud deployment offers several advantages over local deployment, as shown in Table 2-1.

Table 2-1 Cloud Deployment Advantages

Solution Comparison

The mainstream cloud deployment solutions are compared in Table 2-2.

Table 2-2 Cloud Deployment Solution Comparison

Tencent Cloud Lighthouse Deployment (Recommended)

Step 1: Purchase a Server

1. Visit the promotion page:

   https://cloud.tencent.com/act/pro/lighthouse-moltbot

2. Select configuration:

   • Specs: 2 cores, 2GB RAM
   • Bandwidth: 20M
   • Region: Recommended Silicon Valley (overseas regions provide more stable AI model access)
   • Price: 20 CNY/month or 99 CNY/year

3. Identity verification:

   • First-time use requires identity verification
   • Select personal verification
   • Follow the prompts to complete verification

4. Complete purchase:

   • Click "Buy Now"
   • Pay 20 CNY (recommended to buy 1 month first to try)
   • Wait for server creation to complete
   • Optional: Disable auto-renewal

5. Get server information:

   • After purchase, click your avatar → "Internal Messages"
   • Note down:
     • Public IP address
     • Default username (usually lighthouse)
     • Initial password

💡 Free Trial Option (Optional)

Steps:

1. Register for CodeBuddy:

   • International: https://www.codebuddy.ai/promotion/?ref=lweelxalgm (Google/GitHub account)
   • China: https://www.codebuddy.cn/promotion/?ref=7zucxaz7zvqi (phone number)
   • Recommended to register with a new account

2. Claim rewards:

   • After login, click "Practice Reward" → "Claim Now"
   • Get 1 month free usage
   • Accumulate 7 active days to extend by 2 more months (just greet CodeBuddy daily)

3. Reinstall system as OpenClaw:

   • Login to Tencent Cloud console: https://console.cloud.tencent.com/
   • Go to "Lighthouse"
   • Click "Reinstall System"
   • Select "Use Application Template" → "OpenClaw"
   • Select "No Backup Needed", click "Confirm"

4. Next steps:

   • After reinstallation, continue with the steps below

Step 2: Connect to the Server

1. Connect using an SSH client:

   Method 1: Using an SSH Client (Recommended)

   • Download an SSH client (e.g., Xterminal, Termius, FinalShell)
   • Create a new SSH connection:
     • Name: anything you like
     • Address: Public IP address
     • Port: 22
     • Username: lighthouse (or the username from internal messages)
     • Password: the password set during purchase

   Method 2: Using Web Terminal

   • In the Tencent Cloud console, click the instance card
   • Click the "Login" button
   • Open the terminal directly in your browser

2. Verify OpenClaw installation:

   openclaw --version

   If a version number is displayed (e.g., 2026.3.2), OpenClaw is pre-installed successfully.

Step 3: Configure the AI Model

1. Enter Application Management:

   • Click the server card
   • Switch to the "Application Management" tab

2. Select a model:

   • Recommended: Kimi k2.5 (best value for money)
   • You can also choose other domestic AI models

3. Get API Key:

   Kimi k2.5 Configuration (Recommended):

   1. Visit: https://platform.moonshot.cn/
   2. Register and login
   3. Go to "API Management"
   4. Click "Create API Key"
   5. Copy the API Key (format: sk-xxx)

4. Enter configuration:

   • Paste the API Key into the configuration field
   • Click "Save"
   • Wait for the configuration to take effect

Step 4: Test Connection

1. Access WebUI:

   • Use the access URL provided by the console
   • Format: http://YOUR_SERVER_IP:18789/?token=xxx

2. Send a test message:

   Hello, can you hear me?

3. Verify success:
   • If you receive an AI reply, the configuration is successful
   • The model name will be displayed in the top right corner

Volcengine Deployment (Cheaper)

If you're a heavy Lark user, Volcengine is recommended:

1. Visit the promotion page:

   https://www.volcengine.com/activity/clawdbot

2. Price advantage:

   • 9.9 CNY/month
   • 58 CNY/year
   • 10 CNY cheaper than Tencent Cloud

3. Configuration comparison:

   • 2 cores, 2GB RAM (same)
   • 5M bandwidth (Tencent Cloud has 20M)
   • Suitable for Lark users

4. Deployment process:

   • Similar to Tencent Cloud
   • Follow the on-page instructions

Baidu AI Cloud Deployment (0.01 CNY Trial)

Promotion Information

1. Visit the promotion page:

   https://cloud.baidu.com/product/BCC/moltbot.html

2. Promotion rules:

   • First month only 0.01 CNY
   • Requires registration and personal ID verification
   • Limited to one purchase per account

Deployment Steps

Step 1: Purchase a Server

1. Register a Baidu AI Cloud account
2. Complete personal identity verification (bind ID card)
3. Purchase the special LS instance (0.01 CNY/month)
4. If sold out, you can purchase a Lighthouse server at regular price

Step 2: Create an Instance

1. Go to the Lighthouse console
2. Click "Create Instance":
   • Name: auto-generated or custom
   • Password: make sure to remember it, needed for login later
3. Wait for instance creation to complete

Step 3: One-Click OpenClaw Configuration

1. Go to instance details page:

   • Click the instance card
   • Go to the "Application Management" tab

2. Application configuration:

   • Click "One-Click Activation"
   • Click "One-Click Firewall" (open firewall ports)
   • Wait until it shows "Opened"

3. Model configuration:

   • Select a model from the dropdown (e.g., ERNIE Bot, Qianfan)
   • Click "Apply Model Configuration"
   • The system will automatically create a Qianfan API Key and configure it

4. Channel configuration (optional):

   • Supports: Lark, DingTalk, WeCom, QQ
   • Select your preferred messaging tool
   • Follow the prompts to complete configuration (see later chapters for details)

5. Skills configuration (optional):

   • Default: Baidu Search, Baidu Baike
   • Select as needed and click "Apply"
   • More Skills available on the OpenClaw website

Step 4: Access WebUI

1. Click "Get Website URL"
2. Copy the access link
3. Open in your browser to start chatting with OpenClaw

Advantages & Limitations

Advantages:

• ✅ Extremely low price (0.01 CNY for the first month)
• ✅ One-click configuration, no manual installation needed
• ✅ Integrated with Baidu Qianfan AI models
• ✅ Automatic API Key configuration

Limitations:

• ⚠️ First month discount only
• ⚠️ Requires identity verification
• ⚠️ Promotion may sell out

Alibaba Cloud Deployment (Optional)

Alibaba Cloud also offers a one-click OpenClaw deployment solution:

1. Visit the promotion page:

   https://www.aliyun.com/activity/ecs/clawdbot

2. Select Lightweight Application Server:

   • Use the OpenClaw image
   • One-click installation and configuration

3. Price reference:
   • Similar to Tencent Cloud
   • Check the promotion page for specifics

Official Video Tutorials (Highly Recommended)

Tencent Cloud provides detailed video tutorials — following along with the video makes it even easier:

1. Cloud OpenClaw One-Click Deployment with WeCom and QQ Integration

   • Video URL: https://cloud.tencent.com/developer/video/85003
   • Duration: ~10 minutes
   • Content: From purchase to complete configuration

2. Cloud OpenClaw One-Click Deployment with Lark and DingTalk Integration

   • Video URL: https://cloud.tencent.com/developer/video/85055
   • Duration: ~10 minutes
   • Content: Full Lark and DingTalk integration process

Existing Server Deployment

If you already have a Lighthouse server, you can use the AI assistant for conversational deployment:

• Official Tutorial: https://cloud.tencent.com/developer/article/2625605
• Ideal For: Users with existing Lighthouse servers
• Deployment Method: Complete configuration through AI conversation
• Advantage: More flexible, customizable

Cloud Deployment FAQ

Q1: Is cloud deployment secure?

• ✅ Data is stored on your server
• ✅ Only you can access it
• ✅ You can set an access password

Q2: Can I stop it anytime?

• ✅ You can delete the server anytime
• ✅ Billed by usage duration
• ✅ Delete when not needed, no wasted money

Q3: How to access from mobile?

• Through QQ, WeCom, Lark, and other platforms
• See Chapter 9: Multi-Platform Integration

China One-Click Installation (Recommended)

Why Choose the China Version?

The China version offers several localization advantages over the international version, as shown in Table 2-3.

Table 2-3 China Version Advantages

Prerequisites

Required Environment:

• Node.js 22.0.0+ (required)
• pnpm (optional, recommended for source builds)

Recommended Configuration:

• Brave Search API key (for web search)
• Can be configured via openclaw-cn configure --section web

System Requirements:

• macOS: Requires Xcode / Command Line Tools (only CLI + Gateway needs Node.js)
• Windows: WSL2 (Ubuntu) is strongly recommended, native Windows is untested
• Linux: Ubuntu 20.04+, Debian, CentOS

Quick Start

macOS/Linux Installation

# Use the official China installation script
curl -fsSL https://clawd.org.cn/install.sh | bash

Windows Installation

Using PowerShell (Administrator privileges):

# Use the official China installation script
iwr -useb https://clawd.org.cn/install.ps1 | iex

WSL2 Installation Steps:

# 1. Install WSL2
wsl --install

# 2. Restart your computer

# 3. Run the Linux installation command in WSL2
curl -fsSL https://clawd.org.cn/install.sh | bash

Global Installation (Alternative)

If the one-click script fails, you can use npm global installation:

# Using npm
npm install -g openclaw-cn@latest

# Or using pnpm (recommended)
pnpm add -g openclaw-cn@latest

Run the Onboarding Wizard

After installation, run the configuration wizard:

# Run the onboarding wizard and install the background service
openclaw-cn onboard --install-daemon

Configuration Wizard Flow

The wizard will guide you through the following configuration:

1. Select Gateway Mode:

• Local Gateway (recommended): Gateway runs on the local machine
• Remote Gateway: Gateway runs on a server

2. Configure Authentication:

• OpenAI Code (Codex) subscription (OAuth)
• API key (recommended for Anthropic)
• Also supports claude setup-token

3. Select AI Provider:

• Recommended: Kimi, DeepSeek, GLM-4 (domestic models)
• Optional: Claude, GPT (requires API key)

4. Configure Chat Platform (optional):

• WhatsApp: QR code login
• Telegram: Enter Bot Token
• Discord: Enter Bot Token
• Mattermost: Plugin token
• Lark/WeCom/DingTalk: Enter app credentials

5. Install Background Service (recommended):

• macOS: Uses launchd
• Linux: Uses systemd
• WSL2: Uses systemd
• Runtime: Node (recommended, required for WhatsApp/Telegram)

6. Gateway Token:

• The wizard generates a token by default
• Stored in gateway.auth.token
• Generated even on loopback addresses

Authentication Configuration Notes

Authentication Storage Locations (Important):

• Recommended Anthropic path: Set API key (wizard can store it)
• OAuth credentials (legacy import): ~/.openclaw/credentials/oauth.json
• Auth config file: ~/.openclaw/agents/<agentId>/agent/auth-profiles.json

Headless/Server Tips:

• Complete OAuth on a regular machine first
• Then copy oauth.json to the gateway host

Start the Gateway

If you installed the background service, the Gateway should already be running:

# Check Gateway status
openclaw-cn gateway status

Manual Run (Foreground):

# Run in foreground to see real-time logs
openclaw-cn gateway --port 18789 --verbose

Dashboard Access:

Local loopback address: http://127.0.0.1:18789/

If you configured a token, you'll need to paste it into the dashboard settings (stored as connect.params.auth.token).

Quick Verification (2 minutes)

# Check status
openclaw-cn status

# Health check
openclaw-cn health

Pairing + Connecting Chat Interfaces

WhatsApp (QR Code Login)

# Login to WhatsApp
openclaw-cn channels login

Scan the QR code via WhatsApp → Settings → Linked Devices.

Telegram / Discord / Others

The wizard can write tokens/config for you. For manual configuration:

Telegram:

openclaw-cn channels add \
  --channel telegram \
  --token "YOUR_BOT_TOKEN"

Discord:

openclaw-cn channels add \
  --channel discord \
  --token "YOUR_BOT_TOKEN"

DM Security (Pairing Approval)

Default policy: Unknown DMs receive a short code, and messages are not processed until approved.

If the first DM doesn't get a reply, you need to approve the pairing:

# View pairing list
openclaw-cn pairing list whatsapp

# Approve pairing
openclaw-cn pairing approve whatsapp <code>

Running from Source (Development)

If you need to modify OpenClaw itself, you can run from source:

# Clone the repository
git clone https://github.com/clawdbot/clawdbot.git
cd clawdbot

# Install dependencies
pnpm install

# Build UI (automatically installs UI dependencies on first run)
pnpm ui:build

# Build the project
pnpm build

# Run the onboarding wizard
openclaw-cn onboard --install-daemon

If you haven't installed globally yet, you can run commands from the repo via pnpm openclaw-cn ....

Running Gateway from Source:

node dist/entry.js gateway --port 18789 --verbose

End-to-End Verification

In a new terminal, send a test message:

# Send a test message
openclaw-cn message send --target +15555550123 --message "Hello from OpenClaw"

If openclaw-cn health shows "no auth configured", you need to go back to the wizard to set up OAuth/key authentication.

Debugging Tips:

• openclaw-cn status --all: Best read-only debug report
• openclaw-cn health: Requests a health snapshot from the running gateway
• openclaw-cn status --deep: Deep status check

Configuration File Locations

# Main configuration file
~/.openclaw/openclaw.json

# Authentication configuration
~/.openclaw/agents/<agentId>/agent/auth-profiles.json

# OAuth credentials (legacy)
~/.openclaw/credentials/oauth.json

# Log files
~/.openclaw/logs/gateway.log

China Version Special Features

1. Pre-configured Domestic Models:

• Kimi (Moonshot AI)
• DeepSeek
• GLM-4 (Zhipu AI)
• Qwen (Alibaba Cloud)
• ERNIE Bot (Baidu)

2. Domestic Platform Integration:

• Lark (ByteDance)
• WeCom (Tencent)
• DingTalk (Alibaba)
• QQ (Tencent)

3. Optimized Network Configuration:

• Uses domestic mirror sources
• Optimized API access speed
• Supports proxy configuration

Common Issues

Q1: Installation failed?

# Check Node.js version (requires 22+)
node --version

# If version is too low, upgrade with nvm
nvm install 22
nvm use 22

Q2: How to update to the latest version?

# Re-run the installation script
curl -fsSL https://clawd.org.cn/install.sh | bash

Q3: How to uninstall?

# Stop the service
openclaw-cn gateway stop

# Uninstall
npm uninstall -g openclaw-cn

# Delete configuration (optional)
rm -rf ~/.openclaw

Q4: Which systems are supported?

• ✅ macOS 12+
• ✅ Linux (Ubuntu 20.04+, Debian, CentOS)
• ✅ Windows 10/11 (via WSL2)

Q5: Configuration wizard stuck?

# Press Ctrl+C to interrupt

# Check if gateway is running
openclaw-cn gateway status

# Restart gateway and retry
openclaw-cn gateway restart
openclaw-cn onboard

Q6: Health check shows "no auth configured"

You need to configure authentication:

# Re-run the wizard
openclaw-cn onboard

# Or manually configure API key
openclaw-cn configure --section auth

Next Steps (Optional, but Great)

• macOS menu bar app + voice wake: macOS App
• iOS/Android nodes (Canvas/Camera/Voice): Nodes
• Remote access (SSH tunnel / Tailscale service): Remote Access
• Always-on / VPN setup: Tailscale

Next Steps

After installation, you can:

1. Configure AI models (see "API Configuration Guide" below)
2. Connect chat platforms (see Chapter 9: Multi-Platform Integration)
3. Install Skills (see Chapter 8: Skills Extension)
4. Start using (see Chapter 3: Quick Start)