Cloudways Managed AI Agents can be accessed directly from the Cloudways Platform. When you open a supported agent through the Platform, Cloudways can handle the required device access process automatically, so you normally do not need to complete manual device pairing.
However, manual device pairing may be required if you copy the agent URL from the Agent Details section and open it directly in another browser tab, browser, or device instead of accessing the agent through the Cloudways Platform.
Device pairing is a security process that confirms that the browser or device requesting access to your Managed AI Agent is authorized. For agents that require manual approval, you may need to approve the new device before you can access the agent interface.
This guide explains when manual device pairing may be required and how to complete the process for supported Managed AI Agents.
The steps shown in this article use OpenClaw as the example agent.
The exact interface, commands, and setup steps may differ slightly for other Managed AI Agents, such as Hermes.
The existing OpenClaw process confirms that pairing is used to approve a browser, phone, or computer before it can access the agent.
When do you need to manually pair your device?
You normally do not need to manually pair your device when you access a Managed AI Agent through the supported Open Agent workflow on the Cloudways Platform.
Manual pairing may be required when you:
Copy the agent URL from the Agent Details section.
Open the URL directly in another browser tab, browser, or device.
Reach the agent interface directly instead of opening it through the Cloudways Platform.
See a message indicating that the browser or device requires approval.
Need to approve the device using the agent-specific process.
For OpenClaw, the current workflow uses a device pairing request that is approved through the agent server.
The attached process uses the openclaw devices approve <request-id> command after connecting through SSH.
Note:
The manual approval method may vary between Managed AI Agents. The OpenClaw process in this article should not be assumed to apply exactly to Hermes or other agents unless the same workflow is supported.
Before you start
Before starting the manual pairing process, make sure you have:
An active Managed AI Agent deployed on Cloudways.
Access to the Cloudways Platform.
The URL of the agent you want to access.
Any access credentials required by the selected agent.
SSH credentials if the agent-specific manual pairing process requires server access.
A supported terminal or SSH client if SSH access is required.
SSH (Secure Shell) is a secure method of connecting to a server through a command-line interface. For OpenClaw, the Cloudways Platform provides the required Username, IP Address, Port, and Password under the SSH Credentials section.
Important:
Keep your agent access keys, gateway tokens, API keys, SSH credentials, and other authentication details secure. Do not share them publicly or include them in unsecured screenshots or messages.
How manual device pairing works
The manual pairing process generally involves three parts:
Open the Managed AI Agent directly using its agent URL.
Authenticate or generate a device approval request, depending on the selected agent.
Approve the browser or device using the agent-specific method.
After the request is approved, return to the agent interface and reconnect.
For OpenClaw, the approval process uses SSH access to the agent server and a device approval command.
The process supplied in your source confirms that, after approval, the user returns to the browser and reconnects to the OpenClaw Control UI.
Manually pair your device
The exact manual device pairing steps depend on the Managed AI Agent you are using.
Note:
The steps below use OpenClaw-specific controls, credentials, terminology, and commands. Other Managed AI Agents, such as Hermes, may use a slightly different manual access or approval process.
Pairing your Device on Linux/MacOS
Device pairing is OpenClaw's way of confirming that any device trying to reach your agent's dashboard is one you actually control. You approve each new browser, phone, or computer once, and only approved devices can access your agent after that.
It is needed because your OpenClaw agent is reachable over the internet. Without pairing, anyone who knew or guessed the URL could log in and control your agent, your LLM keys, and connected channels like WhatsApp or Slack. Pairing closes that gap.
Feature Update: We are working to automate the device pairing for users that open the agent via Cloudways Platform however the following steps will stay relevant for now.
The first time you access your OpenClaw dashboard from a Linux/MacOS device, you will need to pair that device.
This is a one-time step per device.
Step #1:
Log in your Cloudways Platform
Click on AI Agents on the side bar menu.
Step #2:
Now, click on Open Agent as shown in the image below.
A tab will appear with the following screen.
Step #3:
Copy the Gateway Token/Access Key as shown below.
Step #4:
Paste the key into the OpenClaw login screen. The login screen will display a device ID pending approval.
Click Connect
Step #5:
Now, open a terminal on your computer using SSH credentials
Paste your password in the terminal window as shown below.
Press Enter.
Step #6:
Now, go back to the OpenClaw Gateway Dashboard tab and copy the following text from the instructions to approve the request, as shown below.
To approve the request, paste the command into the terminal window.
Step #7:
Hit enter and the device will be approved
The terminal will confirm the device has been approved.
Step #8:
Once done, click Connect again on the OpenClaw Dashboard and you'll be landed on OpenClaw Dashboard chat
Pairing your Device on Windows
After your OpenClaw agent is deployed, you can access the OpenClaw dashboard directly from the Cloudways Platform.
To open your OpenClaw dashboard:
Log in to your Cloudways Platform.
Navigate to AI Agents.
Select your OpenClaw agent.
Click Open Agent.
Device pairing is now automated when you open your agent from the Cloudways Platform. You no longer need to manually approve your device before accessing the OpenClaw dashboard.
The OpenClaw dashboard allows you to manage your agent configuration, update your LLM provider API key, connect messaging channels, view agent details, and manage agent settings.
Connect to Your OpenClaw Server Using SSH
SSH access is available for users who need advanced server access for configuration, file transfer, or troubleshooting.
SSH stands for Secure Shell. It allows you to securely connect to your OpenClaw server from your computer. The SSH credentials are available inside your OpenClaw agent details page.
Before connecting through SSH, go to your OpenClaw agent page and copy the following details from the SSH Credentials section:
Username
IP Address
Port
Password
Use the steps below based on your operating system.
Connect to SSH on Linux or macOS
Linux and macOS users can connect to the OpenClaw server using the Terminal application.
Open Terminal on your computer.
Run the SSH command using your OpenClaw username, IP address, and port.
Use the following format:
ssh -p <port> <username>@<ip-address>
Example:
ssh -p 22921 [email protected]
Press Enter.
When prompted, enter the SSH password copied from the Cloudways Platform.
Press Enter again.
Once the login is successful, you will be connected to your OpenClaw server through SSH.
Connect to SSH on Windows Using PuTTY
Windows users can use PuTTY to connect to the OpenClaw server through SSH.
Open the PuTTY application on your Windows computer.
In the Host Name (or IP address) field, enter the OpenClaw server IP address.
In the Port field, enter the SSH port shown in the Cloudways Platform.
Make sure the connection type is set to SSH.
Click Open.
If this is your first time connecting to the server, PuTTY may show a security alert for the server host key. Click Accept to continue.
After the terminal opens:
Enter the SSH username copied from the Cloudways Platform.
Press Enter.
Enter the SSH password when prompted.
Press Enter again.
Once the login is successful, you will be connected to your OpenClaw server through SSH.
Go through the following steps in detail to complete this easily.
Step #2 - Copy Your SSH Credentials:
Before connecting through SSH, collect the required SSH credentials from your OpenClaw dashboard.
Under the SSH Credentials section, note down the following details:
Username
IP Address
Port
Password
You will use these details to connect to your OpenClaw server using an SSH client.
Step #3 - Open PuTTY and Configure SSH Connection
Windows users can use PuTTY to establish an SSH connection with the OpenClaw server.
Open the PuTTY application.
Enter your OpenClaw server IP Address in the Host Name (or IP address) field.
Enter the SSH Port available in your OpenClaw dashboard.
Make sure the connection type is set to SSH.
Click Open.
Step #4 - Accept the SSH Host Key:
When connecting to the server for the first time, PuTTY displays a security warning asking you to verify the server host key.
Review the security prompt.
Click Accept to save the host key and continue connecting.
This confirmation is required only the first time you connect to the server.
Step #5 - Log In to Your OpenClaw Server:
After accepting the host key, PuTTY opens the SSH terminal.
Enter the SSH username copied from your OpenClaw dashboard.
Enter the SSH password when prompted.
After successful authentication, you will be connected to your OpenClaw server.
Step #6 - Approve Your Device Pairing Request:
After connecting to your OpenClaw server through SSH, run the following command:
openclaw devices approve <request-id>
Replace <request-id> with the device pairing request ID shown in your OpenClaw Control UI.
Example:
openclaw devices approve cccca975-7001-4505-96e9-8f6db7318ee3
This approves the device request and allows your browser to access the OpenClaw Control UI.
Step #7 - Confirm Device Pairing
After running the approval command successfully, the terminal will display a confirmation message indicating that the device request has been approved.
You can now return to your browser and reconnect to the OpenClaw Control UI.
Important:
Device pairing is required only when accessing the OpenClaw Control UI from a new browser or device. Once the device has been approved, you can continue accessing the Control UI without repeating the pairing process unless a new device requires approval.
What happens after device pairing?
After a device is successfully approved, the authorized browser or device can access the relevant Managed AI Agent interface.
For OpenClaw, the approved device can reconnect to the Control UI after the approval command completes. The existing process also states that pairing is normally required only for a new browser or device.
The behavior may differ for other Managed AI Agents depending on how their authentication and device approval mechanisms work.
Note:
If you use another browser, browser profile, or device, the agent may require a new approval depending on its device access mechanism.
Automatic pairing through the Cloudways Platform
For supported Managed AI Agents, Cloudways simplifies the access process when you open the agent directly from the Platform.
The recommended way to access your agent is:
Cloudways Platform → Managed AI Agents → Select Agent → Open Agent
When the supported automated access flow is used, you do not need to manually complete the pairing procedure described in this article.
The attached OpenClaw workflow already reflects this behavior for Platform-based access, stating that device pairing is automated when the agent is opened from Cloudways.
Important:
Use the manual pairing instructions only when you access the agent directly through its URL or any other external means, and the agent requires device approval.
Troubleshooting manual device pairing
If you cannot access the agent after completing the pairing process:
Confirm that you are approving the correct device or request ID.
Make sure you are using the credentials for the correct Managed AI Agent.
Check whether the agent displays a new approval request.
Return to the agent interface and reconnect after completing the approval.
If you are using OpenClaw, confirm that the device approval command completed successfully before reconnecting.
For OpenClaw, the terminal displays a confirmation after the device request is approved, after which you can return to the browser and reconnect.
Manage Your OpenClaw Agent
Once your OpenClaw agent is active, you can manage your agent configuration and supporting infrastructure.
You can manage:
Skills: Capabilities or tasks your agent can perform.
Nodes: Individual steps or components used within your agent workflows.
Environment Variables: Configuration values used by your agent, such as settings or connection details.
LLM Provider Settings: API keys and provider configurations.
For advanced configuration, file transfers, or troubleshooting, SSH access is available for the underlying server.
SSH (Secure Shell) allows secure command-line access to your server, while SFTP (Secure File Transfer Protocol) allows secure file transfers between your system and the server.
These options are recommended for advanced users or technical teams who need additional control over their OpenClaw environment.
What Cloudways Manages
Cloudways manages the infrastructure layer required to run OpenClaw. This includes droplet provisioning, stack setup, OpenClaw container deployment, OpenClaw installation, SSL/TLS certificate installation and renewal, OpenClaw versioning, and platform-level updates.
SSL/TLS certificates help secure communication between the browser and your OpenClaw instance. Cloudways handles installation and renewal so you do not need to configure certificates manually.
Cloudways also manages the deployment process so you do not need to install Docker, configure the server manually, or manage the runtime stack yourself.
What You Manage
You manage the agent layer. This includes selecting the droplet size and region, connecting your LLM provider API key, securing your API keys and environment variables, defining your agent’s skills and nodes, and controlling how your agent behaves.
You are also responsible for managing your LLM provider usage and costs. Since LLM billing is handled directly by your model provider, you should monitor usage from the provider dashboard and configure any usage limits or billing alerts available from that provider.
Cloudways and Customer Responsibility
The responsibility for OpenClaw on Cloudways is divided between Cloudways and the customer.
Cloudways handles the infrastructure layer. This includes provisioning, deployment, SSL/TLS, platform updates, and the managed setup required to run OpenClaw.
The customer handles the agent layer. This includes selecting the required plan, connecting the LLM provider, managing agent logic, securing keys, and configuring workflows.
Area | Cloudways Responsibility | Customer Responsibility |
Provisioning | Cloudways provisions the droplet in the selected region. | You select the droplet size and region. |
Deployment | Cloudways sets up the stack, deploys the container, and installs OpenClaw. | You connect your LLM provider API key. |
Security | Cloudways installs and renews SSL/TLS certificates. | You secure your API keys and environment variables. |
Updates | Cloudways handles OpenClaw versioning and platform-level updates. | You update agent skills, nodes, and workflow logic. |
Agent Logic | Not managed by Cloudways. | You define the skills, nodes, environment variables, and agent behavior. |
Best practices
Choose a region close to your users to improve response time. If your business has data residency requirements, select a region that matches those requirements.
Keep your LLM provider API key secure. Do not share it publicly or store it in unsafe locations. If you believe your key has been exposed, rotate or regenerate it from your LLM provider account.
Monitor your LLM provider usage regularly. Since LLM costs are billed separately by the provider, reviewing usage can help prevent unexpected charges.
Use backups before making major changes to your agent configuration. This gives you a restore point if you need to recover the previous setup.
Use separate instances for separate agents. This keeps each agent isolated and makes it easier to manage resources, regions, and configurations independently.
Important:
SSH access is mainly recommended for advanced users or technical teams. For general OpenClaw management, you can use the OpenClaw dashboard from the Cloudways Platform.
What we learned
Cloudways can automatically handle device access when you open a supported Managed AI Agent through the Cloudways Platform.
Manual device pairing is mainly required when you access an agent directly using its URL and the agent requests approval for the browser or device.
The exact pairing process depends on the Managed AI Agent. This article uses OpenClaw to demonstrate the current manual workflow, including device approval through SSH.
Other Managed AI Agents, such as Hermes, may use slightly different steps or authentication methods.
That’s it! We hope this article was helpful.