← 返回
AI智能 Key

OpenClaw Setup (ModelWise)

This skill should be used when the user asks to "install OpenClaw", "setup OpenClaw", "configure OpenClaw", "start OpenClaw gateway", "部署 OpenClaw", or needs...
This skill should be used when the user asks to "install OpenClaw", "setup OpenClaw", "configure OpenClaw", "start OpenClaw gateway", "部署 OpenClaw", or needs...
deerleo
AI智能 clawhub v1.0.1 1 版本 100000 Key: 需要
★ 0
Stars
📥 513
下载
💾 6
安装
1
版本
#latest

概述

OpenClaw Installation and Setup Guide

> By ModelWise team - Professional AI Agent Solutions

This skill provides comprehensive guidance for installing, configuring, and running OpenClaw - an open-source personal AI assistant gateway.

Overview

OpenClaw is a multi-channel AI assistant gateway that enables users to use AI assistants on their existing messaging platforms (WhatsApp, Telegram, Slack, Discord, etc.).

Key Features:

  • Local-first architecture
  • 20+ messaging channel integrations
  • Multi-model support (Anthropic, OpenAI, Google, local models)
  • Voice interaction (Voice Wake + Talk Mode)
  • Canvas visualization workspace (A2UI protocol)
  • Skills extension system

Prerequisites

System Requirements

PlatformRequirements
-----------------------
WindowsWindows 10/11, Node.js >= 22.12.0
macOSmacOS 12+, Node.js >= 22.12.0
LinuxUbuntu 20.04+/Debian 11+, Node.js >= 22.12.0

Node.js Installation

Check Node.js version first:

node --version

Windows

# Option 1: Using winget
winget install OpenJS.NodeJS.LTS

# Option 2: Using Chocolatey
choco install nodejs-lts

# Option 3: Download installer
# https://nodejs.org/

macOS / Linux

# Install nvm
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.1/install.sh | bash

# Reload shell
source ~/.zshrc  # or ~/.bashrc

# Install Node.js 22
nvm install 22
nvm use 22
nvm alias default 22

Installation

Step 1: Install OpenClaw CLI

# All platforms
npm install -g openclaw@latest

# Verify installation
openclaw --version

Step 2: Run Configuration Wizard

# Start interactive configuration
openclaw onboard

# Or with daemon installation (recommended)
openclaw onboard --install-daemon

The wizard will guide through:

  1. Model provider selection (Anthropic, OpenAI, Volcengine, etc.)
  2. API key configuration
  3. Channel setup (Telegram, Slack, etc.)
  4. Gateway configuration

Step 3: Configure API Keys

API keys are stored in ~/.openclaw/credentials/ (or %USERPROFILE%\.openclaw\credentials\ on Windows).

# Set Anthropic API key
openclaw credentials set anthropic YOUR_API_KEY

# Set Volcengine API key (for Kimi)
openclaw credentials set volcengine YOUR_API_KEY

# Set OpenAI API key
openclaw credentials set openai YOUR_API_KEY

Step 4: Start Gateway

# Start gateway on default port (18789)
openclaw gateway

# Or specify port
openclaw gateway --port 18789

# Run in background (macOS/Linux)
openclaw gateway --daemon

# Windows: run in new window or use pm2
start /B openclaw gateway

Configuration

Config File Location

PlatformPath
----------------
macOS/Linux~/.openclaw/openclaw.json
Windows%USERPROFILE%\.openclaw\openclaw.json

Essential Configuration

{
  // Gateway settings
  gateway: {
    enabled: true,
    port: 18789,
    mode: "local",
    bind: "0.0.0.0"
  },

  // Agent settings
  agents: {
    defaults: {
      model: "volcengine/kimi-k2.5",
      thinking: {
        enabled: true,
        level: "medium"
      }
    }
  },

  // Telegram channel (example)
  telegram: {
    enabled: true,
    botToken: "YOUR_BOT_TOKEN",
    dmPolicy: "open",
    streaming: "partial"
  }
}

Model Provider Configuration

ProviderModel FormatCredentials Key
----------------------------------------
Anthropicanthropic/claude-opus-4-6anthropic
OpenAIopenai/gpt-5.2openai
Volcenginevolcengine/kimi-k2.5volcengine
Googlegoogle/gemini-2.5-progoogle
Locallocal/llama-3.2N/A

Channel Setup

Telegram Bot Setup

  1. Create Bot:
    • Open @BotFather on Telegram
    • Send /newbot
    • Follow instructions to get bot token
  1. Configure OpenClaw:
  2. openclaw credentials set telegram 7123456789:AAHxxxxxxxx
    openclaw config set telegram.enabled true
    openclaw config set telegram.dmPolicy "open"
    
  1. Restart Gateway:
  2. openclaw gateway restart
    

Security Policies

DM Policy Options:

  • open: Accept messages from anyone (for public bots)
  • pairing: Require pairing code approval (default, more secure)

Common Commands

Diagnostics

# Check system status
openclaw doctor

# View running processes (macOS/Linux)
ps aux | grep openclaw | grep -v grep

# View running processes (Windows PowerShell)
Get-Process | Where-Object {$_.ProcessName -eq "node"}

# Check gateway status
openclaw gateway status

Updates

# Update to latest stable
openclaw update --channel stable

# Update to beta version
openclaw update --channel beta

Maintenance

# Stop gateway (macOS/Linux)
killall openclaw openclaw-gateway

# Stop gateway (Windows PowerShell)
Stop-Process -Name "node" -Force

# View logs
tail -f ~/.openclaw/logs/gateway.log  # macOS/Linux
Get-Content $env:USERPROFILE\.openclaw\logs\gateway.log -Tail 100  # Windows

# Clear sessions
openclaw sessions prune

Directory Structure

PlatformBase Path
---------------------
macOS/Linux~/.openclaw/
Windows%USERPROFILE%\.openclaw\
.openclaw/
├── openclaw.json       # Main configuration
├── credentials/        # API keys (encrypted)
├── sessions/           # Session data
├── logs/               # Log files
├── workspace/          # Working directory
│   └── skills/         # Custom skills
├── agents/             # Agent configurations
├── canvas/             # Canvas data
├── browser/            # Browser profiles
└── memory/             # Memory storage

Troubleshooting

Common Issues

1. Node.js version mismatch:

# Check version
node --version  # Should be >= 22.12.0

# Fix (macOS/Linux)
nvm use 22

# Fix (Windows)
# Download and install from https://nodejs.org/

2. Port already in use:

# macOS/Linux
lsof -i :18789
kill -9 <PID>

# Windows PowerShell
netstat -ano | findstr :18789
taskkill /PID <PID> /F

3. Permission denied:

# macOS/Linux
sudo chown -R $(whoami) ~/.npm

# Windows (run PowerShell as Administrator)
npm cache clean --force

4. Gateway not starting:

# Check logs (macOS/Linux)
cat ~/.openclaw/logs/gateway.log

# Check logs (Windows)
type %USERPROFILE%\.openclaw\logs\gateway.log

# Run diagnostics
openclaw doctor

Reset Configuration

# Backup and reset (macOS/Linux)
cp ~/.openclaw/openclaw.json ~/.openclaw/openclaw.json.bak
rm -rf ~/.openclaw
openclaw onboard

# Backup and reset (Windows PowerShell)
Copy-Item $env:USERPROFILE\.openclaw\openclaw.json $env:USERPROFILE\.openclaw\openclaw.json.bak
Remove-Item -Recurse -Force $env:USERPROFILE\.openclaw
openclaw onboard

Quick Reference

# Installation
npm install -g openclaw@latest

# Configuration
openclaw onboard --install-daemon

# Start/Stop
openclaw gateway              # Start
killall openclaw              # Stop (macOS/Linux)

# Status
openclaw doctor               # Diagnostics
openclaw --version            # Version

# Update
openclaw update --channel stable

Additional Resources

Scripts

Cross-platform installation checkers:

  • scripts/check-installation.sh - Bash script (macOS/Linux/Windows Git Bash)
  • scripts/check-installation.ps1 - PowerShell script (Windows)

Reference Files

  • references/configuration-reference.md - Complete configuration options

Example Files

  • examples/openclaw.json - Full configuration example
  • examples/telegram-setup.md - Telegram channel setup guide

Official Resources

  • Website: https://openclaw.ai
  • Documentation: https://docs.openclaw.ai
  • GitHub: https://github.com/openclaw/openclaw
  • Discord: https://discord.gg/clawd
  • ClawHub (Skills): https://clawhub.com

版本历史

共 1 个版本

  • v1.0.1 当前
    2026-03-30 22:44 安全 安全

安全检测

腾讯云安全 (Keen)

安全,无风险
查看报告

腾讯云安全 (Sanbu)

安全,无风险
查看报告

🔗 相关推荐

ai-intelligence

Proactive Agent

halthelobster
将AI智能体从任务执行者升级为主动预判需求、持续优化的智能伙伴。集成WAL协议、工作缓冲区、自主定时任务及实战验证模式。Hal Stack核心组件 🦞
★ 836 📥 213,118
content-creation

Remotion Video Skill (ModelWise)

deerleo
基于 Remotion 的动画演示视频创作技能,提供丰富的动画组件和视频模板。By ModelWise team.
★ 0 📥 611
ai-intelligence

Self-Improving + Proactive Agent

ivangdavila
自我反思+自我批评+自我学习+自组织记忆。智能体评估自身工作、发现错误并持续改进。
★ 1,358 📥 318,341