Developer Quickstart (Private Alpha)
Every network begins with a small circle of builders. In this Alpha, we’re inviting a handful of pioneers to shape how intelligence grows.
1. Access Approval
ThinkOS is currently in Private Alpha. This means that it's only available to those with a bit of developer experience. When we move to beta it will be installed like a normal application by anyone. To participate, you must be allowlisted.
Approved builders will receive credentials and a private GitHub link.
2. Install ThinkOS on Mac (Private Repository)
Install Docker Desktop Clone the ThinkOS Alpha repository from your private GitHub link:
git clone https://github.com/XXX
cd thinkos-alpha
chmod +x ./setup.sh
./setup.sh
This setup script installs:
The ThinkOS Server (local orchestration layer)
The ThinkOS Browser Extension (for agent interaction)
Basic dependencies for agent management and logging
Once setup completes, start your local node:
thinkos start
Then open the dashboard at:
👉 http://localhost:
8082
3. Connect Your Openrouter Key (in the Server Interface)
Agents are created directly inside the ThinkOS Server Interface. Create an Open Router key. Once logged in, select “Create” from the control panel on the left and select "Secret". Add your Open Router key and name the secret.
Select the Open Router tool from the green tool nodes, and click "Configure Tool", select the secret key you'd like to use and the model from the drop down.
Next select the "Default Chat Assistant" and select the Open Router model you'd like to use.
4. Browser Extension
Your Think Agent is now automatically communicating with the Think Agent extension inside the browser. If you run into any issues with the browser—go to "Manage Extensions" in your browser and hit the restart icon which should reload the extension.
Principle in Action
Alpha participants are not just testers; they are co-architects of an ecosystem. Every bug report, model tweak, and feature suggestion compounds value for everyone who comes after. We grow this network the same way we grow intelligence itself — iteratively, cooperatively, and with purpose.
Last updated