Deploy OpenClaw: Open-Source AI Personal Assistant from Installation to QQ Bot Integration

OpenClaw is an open-source AI personal assistant that runs on your local machine and interacts with you through various chat applications. It supports multiple AI model providers and features a rich plugin and skill system, making it easy to build your own AI assistant.

Website: https://openclaw.ai
Documentation: https://docs.openclaw.ai

What is OpenClaw?#

OpenClaw is a local-first open-source AI personal assistant framework. Key features include:

Multi-platform chat integration: Supports QQ, Telegram, Discord, WeChat, and other popular chat platforms
Model agnostic: Works with Anthropic, OpenAI, DeepSeek, Moonshot, and many other AI models
Plugin/skill system: Extend functionality through plugins with support for custom skills
Local-first: Data stored locally to protect your privacy
Visual dashboard: Built-in management dashboard

Supported chat channels:

QQ (via QQ Open Platform)
Telegram
Discord
WeChat (Enterprise WeChat)
Slack
Command Line (CLI)

Supported AI model providers:

Provider	Recommended Model	Notes
Anthropic	Claude 4 Sonnet	Strong overall capabilities
OpenAI	GPT-4o	Great multimodal support
OpenRouter	Various models	Aggregation platform, choose as needed
DeepSeek	DeepSeek-V3	Cost-effective
Moonshot (Kimi)	Kimi 2.5	Optimized for Chinese, long context
Local models	Ollama / LM Studio	Fully offline operation

Prerequisites#

OpenClaw runs on Node.js and requires Node.js 22+.

It’s recommended to use nvm to manage your Node.js version:

1
# Install nvm
2
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.1/install.sh | bash
3

4
# Install Node.js 22
5
nvm install 22
6
nvm use 22
7

8
# Verify version
9
node --version  # Should show v22.x.x

NOTE
For users in China experiencing slow npm downloads, you can use the npmmirror registry:
Terminal window
1
npm config set registry https://registry.npmmirror.com

Step 1: Install OpenClaw#

Install OpenClaw globally via npm:

1
npm install -g openclaw

Verify the installation:

1
openclaw --version

Step 2: Run the Onboarding Wizard#

After installation, run the onboarding wizard to configure OpenClaw:

1
openclaw onboard

The wizard will guide you through the following steps.

Security Notice#

On first run, a security notice will appear informing you that OpenClaw will start a Gateway service locally. Confirm and select Yes to continue.

Choose Onboarding Mode#

The wizard offers two modes:

QuickStart (Recommended): Quick setup, suitable for most users
Advanced: Advanced configuration with more customization options

It’s recommended to choose QuickStart mode to get started quickly.

Configure AI Model#

Select your AI model provider and enter the corresponding API Key.

Provider	API Key URL	Notes
Anthropic	https://console.anthropic.com	Claude series models
OpenAI	https://platform.openai.com	GPT series models
OpenRouter	https://openrouter.ai	Aggregates multiple models
DeepSeek	https://platform.deepseek.com	Cost-effective Chinese provider
Moonshot	https://platform.moonshot.cn	Kimi series models
Local models	Ollama / LM Studio	No API Key needed

Here we use Kimi 2.5 as an example — just get your API Key from the Moonshot platform and enter it. The configuration process is similar for other providers.

TIP
For detailed configuration instructions for each provider, refer to the OpenClaw official documentation: https://docs.openclaw.ai/models

Select Chat Channels#

Choose the chat platforms you want to integrate. You can select multiple channels and add more later from the Dashboard.

NOTE
The QQ channel requires additional configuration steps (registering on QQ Open Platform, creating a bot, etc.), which will be covered in detail in the next section. You can skip QQ during initialization and set it up later.

Configure Skills#

Choose skills based on your needs. If you’re on Linux/macOS, it’s recommended to install brew.

NOTE
These API Keys are all optional and used for specific skills. If you don’t have them, just select No for all and skip — you can add them later via openclaw configure at any time.

Confirm UI#

This is the final step.

Option	Description
Hatch in TUI (recommended)	Enter the interactive TUI interface directly in the terminal to chat with the Bot and set up its persona. Recommended.
Open the Web UI	Open the browser-based Web dashboard to complete the initial configuration.
Do this later	Skip the initialization process and set it up later.

After selecting Hatch in TUI, you’ll enter the terminal chat interface:

1
openclaw tui - ws://127.0.0.1:18789 - agent main - session main

The Bot will send “Wake up, my friend!” as its first message. You can start chatting with it, telling it your needs and preferences — the more detailed your description, the better the subsequent experience.

To exit TUI: Press Ctrl+C. The Bot’s Gateway service will continue running in the background.

Complete Installation#

After the wizard finishes, OpenClaw will automatically start the Gateway service.

Verify the running status:

1
openclaw status

Open the visual management dashboard:

1
openclaw dashboard

The default dashboard address is http://localhost:3210, where you can view and manage your AI assistant.

Step 3: QQ Bot Integration#

To integrate OpenClaw with a QQ bot, follow these steps.

Register on QQ Open Platform#

Visit QQ Open Platform
Scan the QR code with your mobile QQ to log in or register a developer account
Complete developer verification

Create a QQ Bot#

After logging in, go to the QQ Bot Management Page
Click Create Bot
Fill in the bot’s basic information (name, avatar, description, etc.)
After creation, obtain the AppID and AppSecret from the bot’s detail page

IMPORTANT
The AppID and AppSecret are the bot’s core credentials. Keep them safe and never expose them in public repositories or share them with others.

Install the QQ Bot Plugin and Configure#

Install the OpenClaw QQ Bot plugin:

1
openclaw plugins install @sliverp/qqbot@latest
2

3
openclaw channels add --channel qqbot --token ""
4

5
openclaw gateway restart

Usage#

Security Considerations#

WARNING
The OpenClaw Gateway service listens on a local port by default. Please pay attention to the following security recommendations:

Do not expose the Gateway port directly to the public internet to prevent unauthorized access
For remote access, use a reverse proxy (such as Nginx or Caddy) with HTTPS encryption
Regularly update OpenClaw to the latest version to get security fixes
Do not commit API Keys or other sensitive information to version control systems

Common Commands#

Command	Description
`openclaw init`	Run the initialization wizard
`openclaw status`	Check running status
`openclaw dashboard`	Open the management dashboard
`openclaw gateway start`	Start the Gateway
`openclaw gateway stop`	Stop the Gateway
`openclaw gateway restart`	Restart the Gateway
`openclaw plugins install <name>`	Install a plugin
`openclaw plugins list`	List installed plugins
`openclaw skills list`	List available skills
`openclaw config edit`	Edit the configuration file
`openclaw update`	Update to the latest version

Summary#

In this guide, we completed the following steps:

Install OpenClaw: Global installation via npm
Initial configuration: Run the wizard to configure AI models and chat channels
QQ Bot integration: Register on QQ Open Platform, create a bot, install the plugin, and configure

OpenClaw’s plugin and skill system is very flexible. We recommend further exploration:

Install more plugins to extend functionality
Try different AI models to compare performance
Create custom skills to meet your specific needs

Related Links:

OpenClaw Website: https://openclaw.ai
OpenClaw Documentation: https://docs.openclaw.ai
OpenClaw GitHub: https://github.com/openclaw/openclaw
QQ Open Platform: https://q.qq.com