OpenClaw GitHub Installation Guide: Clone, Setup and Run OpenClaw on a VPS 

Key highlights 

• Learn how to install OpenClaw from GitHub using a straightforward clone repository guide that walks you through each step without confusion. 

• Understand the most common OpenClaw installation errors so you can identify what went wrong and apply the right fix the first time. 

• Explore the OpenClaw GitHub page to locate the source files, repository structure and documentation you need before starting setup. 

• Uncover the key configuration steps that follow a successful repository clone to make sure OpenClaw is fully operational on your server. 

Developers increasingly want to run AI systems within their own infrastructure rather than rely on SaaS tools that process data externally. OpenClaw is an open-source platform that lets teams build and run AI agents, automate workflows and orchestrate AI tasks within their own infrastructure. 

In this OpenClaw GitHub guide, you’ll learn how to install OpenClaw from GitHub, configure the required tools and run the platform locally or on a server. The guide also includes an OpenClaw clone repository guide, step-by-step installation instructions and solutions for common OpenClaw installation errors. 

By the end, you’ll understand how to set up OpenClaw correctly, troubleshoot issues and run a self-hosted AI automation system with full control over data, integrations and execution. 

What is OpenClaw on GitHub? 

OpenClaw is an open-source AI automation platform available on GitHub that allows developers to build, run and manage AI agents within their own infrastructure. Instead of relying on external SaaS AI tools, OpenClaw lets teams create self-hosted AI systems that integrate directly with internal workflows, APIs and applications. 

By running OpenClaw locally or on a VPS server, developers gain full control over their data, AI models and automation logic. This makes it useful for organizations that want to move from simple AI experiments to reliable, production-ready AI workflows running inside their own environment.  

Also read: What Is OpenClaw AI? The Open-Source AI Agent Explained 

1. Overview of the OpenClaw repository 

The OpenClaw GitHub repository contains the complete source code required to install and run the platform. It includes the OpenClaw CLI, backend services and a web dashboard used to manage AI agents and workflows. 

Developers typically clone the repository, install dependencies and build the project locally before launching the OpenClaw gateway. The repository also contains configuration files, scripts and documentation that guide the installation process. 

2. What developers use OpenClaw for 

Developers use OpenClaw to build AI-powered automation systems that can operate across multiple tools and services. 

Common use cases include: 

• Creating internal AI assistants 

• Automating workflows and repetitive tasks 

• Connecting AI models to APIs and tools 

• Running autonomous AI agents 

• Integrating AI into messaging platforms like Slack or Discord 

Because OpenClaw supports tool calling, workflow pipelines and multi-step reasoning, it enables teams to create repeatable AI processes instead of one-off prompts. 

Also read: VPS With High RAM: Boost Speed, Stability and Scalability for Your 2026 Projects 

3. System requirements before installation 

Before installing OpenClaw from GitHub, make sure your environment meets the basic system requirements. 

Recommended minimum specifications: 

Component	Requirement
CPU	2–4 cores
RAM	8 GB recommended
Storage	10 GB available space
Network	Internet access for dependencies

You also need a supported operating system, Node.js v22 or higher and the pnpm package manager installed before starting the setup process. 

Also read: VPS Hosting for Developers: Full Control, Speed and CI/CD Support  

What are the prerequisites for installing OpenClaw? 

Before you install OpenClaw from GitHub, make sure your environment includes the required operating system support and development tools. Preparing these prerequisites helps avoid dependency issues and reduces the chance of common installation errors during setup. 

OpenClaw relies on modern JavaScript tooling and terminal-based commands, so developers typically install it in a Unix-like environment or through Windows Subsystem for Linux. 

Supported operating systems 

OpenClaw can run on several modern operating systems used by developers and server environments. This flexibility allows teams to install and test the platform locally during development or deploy it on servers for production workloads. 

1. macOS 

macOS supports OpenClaw installation through the built-in Terminal application. Developers can install dependencies such as Node.js and pnpm using package managers like Homebrew. 

Because macOS uses a Unix-based environment, it works well for cloning repositories, installing packages and running development servers. 

2. Linux 

Linux is one of the most common environments for running OpenClaw, especially when deploying on servers or VPS infrastructure. Popular distributions such as Ubuntu, Debian and Fedora provide strong compatibility with Node.js and development tools. 

Linux systems also offer better control over permissions, networking and automation, making them ideal for production deployments. 

Also read: Linux Hosting: Introducing a Reliable, Cost-Effective Hosting Solution 

3. Windows (WSL2 recommended) 

OpenClaw can run on Windows, but most developers use Windows Subsystem for Linux (WSL2) to ensure compatibility with Linux-based tooling. 

WSL2 provides a Linux terminal environment inside Windows, allowing you to run commands like cloning repositories, installing packages and managing development workflows without compatibility issues. Once your operating system environment is ready, the next step is installing the required development tools needed to run OpenClaw. 

Required software 

In addition to a supported operating system, several tools must be installed before setting up OpenClaw. These tools ensure the platform can install dependencies, compile the source code and run the gateway services correctly. 

Tool	Requirement	Description	Command / action
Node.js	Version 22 or higher	Required to run OpenClaw backend services and CLI tools.	Check version: node -v
pnpm	Latest version recommended	Used to manage project dependencies and packages efficiently.	Install globally: npm install -g pnpm
Terminal environment	Bash, Zsh, PowerShell or Linux shell	Required to run installation commands, manage dependencies and launch OpenClaw.	Examples: Bash, Zsh, Windows PowerShell, Windows Terminal

1. Node.js version 22 or higher 

OpenClaw requires Node.js version 22 or later to run its backend services and CLI tools. 

You can check your Node.js version using: 

node -v  

If Node.js is not installed or the version is outdated, download the latest version from the official Node.js website. 

2. pnpm package manager 

OpenClaw uses pnpm to manage project dependencies and packages. 

Install pnpm globally with: 

npm install -g pnpm  

Using pnpm improves dependency installation speed and helps maintain consistent package management across development environments. 

3. Bash or PowerShell terminal 

You will need a command-line interface to run installation commands and manage the OpenClaw environment. 

Common terminal options include: 

• Bash 

• Zsh 

• Windows PowerShell 

• Windows Terminal 

• Linux shell environments 

These terminals allow you to clone the repository, install dependencies, build the project and launch OpenClaw. Once your environment and tools are ready, you can proceed to install OpenClaw directly from its GitHub source code. 

How do you install OpenClaw from the GitHub source code? 

Installing OpenClaw directly from the GitHub source code is the recommended approach if you want to customize the platform, contribute to the project or build custom AI workflows. This method gives developers full access to the codebase and configuration files, while allowing them to run OpenClaw locally or on a server. 

The steps below explain how to install OpenClaw from GitHub, build the application and launch the gateway. 

Step 1: Clone the OpenClaw GitHub repository 

Start by opening your terminal and cloning the official repository to your local machine. Then move into the project directory. 

git clone https://github.com/openclaw/openclaw.git 
cd openclaw  

This step downloads the full source code and begins the OpenClaw clone repository guide, which prepares the project for installation. 

Step 2: Install dependencies and build the application 

OpenClaw uses pnpm to manage project dependencies and packages. 

If pnpm is not installed yet, install it globally: 

npm install -g pnpm  

Install project dependencies 

Next, install all required packages for the repository: 

pnpm install  

This command downloads and installs all dependencies needed for the OpenClaw backend services and dashboard. 

Build backend and dashboard UI 

After installing dependencies, build the backend gateway and web interface. 

pnpm build 
pnpm ui:build  

These commands compile the TypeScript code for both the backend services and the dashboard UI. 

Step 3: Link OpenClaw globally and run onboarding 

To make the OpenClaw command accessible from anywhere on your system, link the package globally and start the onboarding process. 

pnpm link --global 
openclaw onboard --install-daemon  

The onboarding wizard helps configure: 

• API keys for AI models 

• background services 

• system configuration 

• workspace settings 

Step 4: Launch OpenClaw in development mode 

If you’re developing custom skills or modifying the core code, you can run OpenClaw in development mode with hot reloading. 

pnpm gateway:watch  

This command launches the gateway and automatically reloads changes whenever the code is updated, speeding development and testing. After starting the gateway, the next step is to confirm that the installation was successful and that all OpenClaw services are running correctly. 

How can you verify that OpenClaw is installed correctly? 

After installing OpenClaw from GitHub, verify the CLI, gateway, services and dashboard work correctly before connecting integrations or messaging channels. This helps identify configuration or dependency issues early. 

Check the gateway health and status 

The OpenClaw gateway is the core service that processes AI requests and manages agent communication across different channels. Confirming gateway health ensures the RPC server (Remote Procedure Call server that processes AI requests) is running correctly on port 18789. 

1. Run diagnostic command 

OpenClaw provides a built-in diagnostic command that scans your system for configuration or environment issues. 

Run: 

openclaw doctor  

This command checks: 

• Node.js version compatibility 

• configuration file validity (~/.openclaw/openclaw.json) 

• open ports and network settings 

• API key configuration 

• system permissions 

Results appear in color-coded output: 

• Green: configuration is correct 

• Red: an issue needs attention 

Example output: 

✓ CLI: v1.2.3   
✗ Config: Missing ANTHROPIC_API_KEY → openclaw config set anthropic.apiKey sk-...   
✓ Gateway: Running on 127.0.0.1:18789  

If problems appear, you can run: 

openclaw doctor --fix  

This command attempts to automatically resolve common issues such as missing daemon services or invalid configuration files. 

2. Check service status 

To quickly review the system status, run: 

openclaw status  

This command displays: 

• gateway uptime 

• active sessions 

• connected channels 

For more detailed diagnostics, run: 

openclaw status --all  

This extended check shows memory usage, running agents and deeper service diagnostics. You can also test the gateway directly using: 

openclaw gateway status  

On Linux systems using systemd, you can verify the persistent daemon service with: 

systemctl status openclaw-gateway  

If the installation was completed with openclaw onboard –install-daemon, the service should appear as active (running). Once the gateway service is confirmed to be running, you can access the OpenClaw dashboard to manage and monitor the system. 

How do you access the OpenClaw dashboard? 

The OpenClaw dashboard provides a visual interface where you can manage agents, monitor logs, configure integrations and view system activity in real time. It also helps verify that your OpenClaw installation is running correctly and that all services are functioning as expected. 

1. Default dashboard URL 

Once the gateway is running, open your browser and navigate to: 

http://127.0.0.1:18789  

or 

http://localhost:18789  

The dashboard includes sections such as: 

• Channels 

• Agents 

• Configuration 

• Logs 

You can also launch the dashboard directly from the terminal: 

openclaw dashboard  

If a firewall blocks the default port, change it with: 

openclaw config set gateway.port 8080  

Then restart the gateway: 

openclaw restart  

Security note: If the dashboard becomes accessible externally, restrict access using firewall rules or a VPN to prevent unauthorized access. 

Once the dashboard loads successfully, it confirms that your OpenClaw installation is working correctly and ready for AI automation workflows. However, some developers may still encounter setup issues during installation or configuration. 

What are the most common OpenClaw GitHub installation issues? 

Even when following the official steps to install OpenClaw from GitHub, developers may encounter setup problems related to dependencies, configuration files or service startups. Identifying these issues early helps ensure the gateway and services run correctly. 

Below are some common OpenClaw installation errors and practical troubleshooting steps to resolve them. 

1. Installation script not running 

One of the most common issues occurs when installation commands fail to execute properly. 

This can happen due to: 

• Outdated Node.js versions 

• Missing pnpm package manager 

• Incorrect shell environment 

• Insufficient system permissions 

How to fix it 

• Verify your Node.js version: 

node -v  

OpenClaw requires Node.js v22 or later. 

• Ensure pnpm is installed globally: 

npm install -g pnpm  

• Run commands from a supported terminal, such as Bash, Zsh or PowerShell. 

If installation scripts still fail, reinstall dependencies and restart the setup process. 

2. Dependency or pnpm errors 

Dependency issues are another frequent cause of installation problems. 

Typical symptoms include: 

• failed package installation 

• corrupted dependency cache 

• version conflicts between packages 

OpenClaw error fixes 

Clear the pnpm store and reinstall dependencies: 

pnpm store prune 
pnpm install  

If the problem persists, remove the lock file and reinstall packages: 

rm pnpm-lock.yaml 
pnpm install  

This rebuilds the dependency tree and often resolves installation conflicts. 

3. API key configuration problems 

OpenClaw requires API keys for supported AI model providers. If these keys are missing or incorrectly configured, the gateway may fail to process requests. 

Common signs include: 

• authentication errors 

• failed AI responses 

• warnings during startup 

How to fix it 

Add the required API keys through the configuration command: 

openclaw config set anthropic.apiKey YOUR_API_KEY  

You can also verify configuration settings using: 

openclaw config show  

Ensuring API keys are correctly configured is essential for running AI agents and automation workflows. 

4. Gateway service not starting 

In some cases, the OpenClaw gateway fails to start after installation. 

This issue is typically caused by: 

• port conflicts 

• missing configuration values 

• failed background services 

Troubleshooting steps 

Check the gateway status: 

openclaw gateway status  

Review system logs for errors and confirm the default port (18789) is not already in use. 

If needed, restart the service: 

openclaw restart  

These OpenClaw error fixes usually resolve gateway startup problems and restore normal system operation.

What security best practices should you follow when running OpenClaw? 

Because OpenClaw can run AI agents that interact with APIs, tools and system resources, securing your environment is essential. A misconfigured setup could expose sensitive data, allow unauthorized access or let agents execute unsafe commands. 

Following a few core security practices helps protect your system while running OpenClaw locally or on a server. 

1. Protect the OpenClaw dashboard port 

The OpenClaw dashboard provides full administrative access to agents, integrations and configuration settings. By default, it runs on port 18789, so exposing it publicly can create a serious security risk. 

To prevent unauthorized access, bind the gateway only to localhost in the configuration file: 

"gateway": { 
  "host": "127.0.0.1", 
  "port": 18789 
}  

This ensures the dashboard is accessible only from the local machine. 

If you need remote access, use a secure tunnel such as: 

• SSH tunneling 

• Tailscale 

• Cloudflare Tunnel 

For example: 

ssh -L 18789:localhost:18789 user@server  

Avoid directly forwarding port 18789 through your router or firewall. You can also block external access with firewall rules such as: 

ufw deny 18789  

To monitor active connections, run: 

netstat -tlnp | grep 18789  

2. Restrict AI command execution 

OpenClaw agents can run shell commands, call APIs and interact with external services. Without proper controls, this could allow unsafe actions. 

You can restrict command execution by configuring allowlists and sandboxing in the security settings: 

"security": { 
  "sandbox": true, 
  "tools": { 
    "allowlist": ["calc", "curl", "git status"], 
    "deny": ["rm", "sudo", "dd"], 
    "maxTimeout": 30 
  }, 
  "workspaceRoot": "/safe/readonly/path" 
}  

These settings limit what commands agents can execute and prevent access to critical system operations. 

Additional security practices include: 

• disabling untrusted plugins or integrations 

• manually auditing custom skills or scripts 

• enabling mention-only mode so agents run only when triggered with commands like @openclaw 

Using modern AI models with stronger prompt-injection protection can also reduce the risk of malicious prompts affecting agent behavior. 

3. Secure API keys and credentials 

API keys stored in the OpenClaw configuration file allow access to AI model providers and external services. If these keys are exposed, attackers could run unauthorized workloads or consume your API credits. 

To protect credentials, restrict file permissions: 

chmod 600 ~/.openclaw/openclaw.json  

You should also ensure the configuration directory belongs to the correct user: 

chown $USER ~/.openclaw/  

A safer approach is to store API keys in environment variables or a secrets manager rather than directly in configuration files. 

Example .env configuration: 

ANTHROPIC_API_KEY=sk-... 
OPENAI_API_KEY=sk-...  

Additional best practices include: 

• Setting spending limits on AI provider accounts 

• Rotating API keys regularly 

• Creating restricted API keys with limited permissions 

• Avoiding committing configuration files to Git repositories 

Following these security practices helps prevent common risks such as dashboard hijacking, unauthorized command execution and API credential exposure while running OpenClaw. Once your environment is secured, it’s also important to keep OpenClaw updated to maintain stability and receive the latest improvements. 

How do you update OpenClaw after installation? 

OpenClaw is actively developed, so new updates frequently introduce bug fixes, security improvements and new features. Keeping your installation updated helps improve stability and ensures compatibility with the latest AI models, integrations and workflows. 

The update process depends on how OpenClaw was originally installed. If you installed it through the CLI, you could update it using package manager commands. If you installed it directly from GitHub source code, updates require pulling the latest repository changes and rebuilding the project. 

1. Updating CLI installations 

If OpenClaw was installed globally using the CLI, updating it is straightforward. 

Run the following command to update the CLI package: 

pnpm update -g openclaw  

After updating, verify the installed version: 

openclaw --version  

If the gateway service is running as a daemon, restart it to apply the update: 

openclaw restart  

This ensures the updated version of the CLI and services is running correctly. 

2. Updating source installations from GitHub 

If you installed OpenClaw directly from its GitHub repository, updating requires pulling the latest changes and rebuilding the application. 

Navigate to the project directory and run: 

git pull  

Next, update dependencies and rebuild the project: 

pnpm install 
pnpm build 
pnpm ui:build  

Finally, restart the OpenClaw gateway to apply the updates. 

Updating regularly helps maintain system stability, reduces compatibility issues and ensures your OpenClaw environment benefits from the latest improvements. Once updated, you can consider deploying OpenClaw on a VPS to run AI agents continuously in a production environment. 

Also read: VPS with Unlimited Bandwidth for High-Traffic Websites 

Deploying OpenClaw on Bluehost VPS 

While OpenClaw can run on a local machine for testing, most production deployments run on a VPS so AI agents can operate continuously and integrate into internal systems. and integrate them into production workflows. Hosting OpenClaw on a server ensures reliable uptime, stronger security controls and the ability to scale AI workloads as usage grows. 

Using a self-managed VPS allows organizations to run private AI infrastructure without depending on external SaaS tools, keeping sensitive data and automation logic within their own environment.  

Why use VPS for OpenClaw 

Running OpenClaw on a VPS offers several advantages over local installations.  

1. Dedicated infrastructure 

A VPS gives OpenClaw dedicated compute resources such as CPU, RAM and storage. This allows AI agents and workflows to run consistently without being limited by local machine performance. 

2. Better security 

Self-hosting OpenClaw on your own server keeps prompts, API keys and internal data inside your infrastructure. This approach helps organizations avoid sending sensitive information to third-party AI platforms and maintain full control over their systems.  

3. More scalability 

As AI workloads grow, a VPS can be upgraded with more CPU cores, memory and storage. This flexibility allows teams to run multiple AI agents, automation workflows and integrations without performance bottlenecks. 

Deploying OpenClaw with one-click installation 

Bluehost VPS provides a simplified way to deploy OpenClaw without manually configuring every dependency. This allows developers to launch a self-hosted AI automation environment quickly while maintaining full control over infrastructure and data. 

1. Set up VPS 

Start by launching a self-managed VPS instance with the required resources for OpenClaw. Choose a plan that provides sufficient CPU, memory and storage depending on your expected AI workloads. 

2. Install OpenClaw 

Once the VPS environment is ready, install OpenClaw using the one-click deployment option or by running the installation commands from the GitHub repository. This installs the gateway, dashboard and required services needed to run AI agents. 

3. Configure integrations 

After installation, configure integrations such as: 

• AI model providers like OpenAI or Anthropic 

• messaging platforms such as Slack or Discord 

• automation tools and APIs 

These integrations allow OpenClaw agents to interact with external services, trigger workflows and execute tasks automatically across your systems.  

Deploy OpenClaw on Bluehost Self-Managed VPS to run AI agents continuously with dedicated resources, stronger security and scalable infrastructure. Get in touch with our Bluehost customer support for help setting up and managing your AI automation workflows. 

Final thoughts 

OpenClaw makes it easier for developers and technical teams to build and run self-hosted AI agents and automation workflows within their own infrastructure. By installing OpenClaw from GitHub, you gain full control over AI models, integrations and system configurations while avoiding dependency on external SaaS tools. 

Once installed, you can customize workflows, connect APIs and deploy AI agents that automate tasks across your systems. For production environments, running OpenClaw on a VPS ensures reliable uptime, stronger security and the ability to scale as your AI workloads grow. 

If you plan to run AI agents continuously, consider deploying OpenClaw on our Bluehost OpenClaw to get dedicated resources, better performance and full infrastructure control for your AI automation stack. 

FAQs