OpenClaw is a self-hosted AI assistant that works with your devices and interacts with the apps you already use—like WhatsApp, Telegram, Slack, Discord, and more. It can answer questions, automate tasks, interact with your files and services, and talk or listen to supported devices, all while keeping you in control of your data.
Rather than just another chatbot, OpenClaw serves as a true personal assistant that fits into your daily commute. In just a few months, this open source project has grown in popularity, surpassing 150,000+ stars on GitHub. In this article, we will go through how to get started with OpenClaw and connect it to WhatsApp.
What can OpenClaw do?
OpenClaw is designed to fit seamlessly into your existing digital life. It meets with 50+ integrationsallows you to chat with your assistant from apps like WhatsApp, Telegram, Slack, or Discord, while controlling and automating tasks from your desktop. You can use cloud or local AI models of your choice, manage notes and tasks, control music and smart home devices, trigger automations, and interact with files, browsers, and APIs—all from one proprietary assistant.
Outside of chat, OpenClaw serves as a powerful hub for automation and productivity. It works with popular tools like Notion, Obsidian, GitHub, Spotify, Gmail, and Home Assistants, supports voice interaction and live virtual Canvas, and works across macOS, Windows, Linux, iOS, and Android. Whether you’re scheduling jobs, managing devices, generating content, or automating workflows, OpenClaw brings everything together under one private, scalable AI assistant.
Includes OpenClaw
You can head over to openclaw.ai to access the code and follow the quick start guide. OpenClaw supports macOS, Windows, and Linux, and provides a simple one-liner that installs Node.js and all the dependencies you need:
curl -fsSL -o install.cmd && install.cmd && del install.cmdAfter running the command, OpenClaw will guide you through the onboarding process. During setup, you will see security-related warnings explaining that the assistant can access local files and perform actions. This is expected behavior—as OpenClaw is designed to automate, it also highlights the importance of being vigilant about information and permissions.
Preparing for the LLM
Once the setup is complete, the next step is to choose an LLM provider. OpenClaw supports many providers, including OpenAI, Google, Anthropic, Minimax, and others.
After selecting your provider, you will be asked to enter the corresponding API key. Once the key is verified, you can select the specific model you want to use. For this setup, we will be using GPT-5.1.
Adding Skills
During configuration, OpenClaw also allows you to add capabilities, which define what the agent can do beyond basic chat. OpenClaw uses skill folders associated with AgentSkills to teach the agent how to work with different tools and services.
Each skill resides in its own directory and includes a SKILL.md file with YAML frontmatter and implementation instructions. By default, OpenClaw loads the bundled capabilities with any local release, and filters them initially based on your location, configuration, and available binaries.
OpenClaw also supports ClawHub, a simple skill registry. If enabled, the agent can automatically search for relevant skills and insert them where needed.
Another popular option is to simply search for the skill you need, copy the command provided, and ask the agent to use it. Once used, a new skill is added and immediately available in OpenClaw.
Configuring a Chat Channel
The last step is to configure the channel where you want to use the agent. In this application, we will use WhatsApp. During setup, OpenClaw will ask for your phone number and display a QR code. Scanning this QR code links your WhatsApp account to OpenClaw.
Once connected, you can send a message to OpenClaw on WhatsApp—or any other supported chat app—and it will reply directly to the same conversation.
Once setup is complete, OpenClaw will open a local web page in your browser with a unique gateway token. Be sure to keep this token safe and handy, as it will be needed later.
It uses the OpenClaw Gateway
Next, we’ll start the OpenClaw Gateway, which acts as the OpenClaw control plane. Gateway uses a WebSocket server that manages channels, nodes, sessions, and hooks.
To start the Gateway, run the following command:
If the Gateway is running, refresh the local web page that displayed the token. This will open the OpenClaw Gateway dashboard.
From the dashboard, navigate to the Overview section and enter the Gateway token you saved earlier to complete the connection.
Once this is done, you can start using OpenClaw from the Gateway dashboard chat or by sending the bot directly to WhatsApp.
Note that OpenClaw responds to messages sent to you on WhatsApp, so be sure to chat with your number when testing the setup.
I am a Civil Engineering Graduate (2022) from Jamia Millia Islamia, New Delhi, and I am very interested in Data Science, especially Neural Networks and its application in various fields.
