Getting Started with Coder Workspaces

Welcome to Coder! This guide will help you create your first workspace and connect to it using various tools.

Creating a Workspace

Visit https://coder.dreamlab.ucsb.edu.
Click UCSB Login and log in using UCSB SSO (via Google).
Click the New Workspace button.
On the New Workspace page:
- Give the workspace a name (required).
- Check I understand the usage policies (required).
- Select additional software and workspace options to enable (see below).
- Click Create Workspace.
- It may take a few minutes for the workspace to boot.

New Workspace Options

When creating a workspace, you can configure the following options:

RStudio: Enables the RStudio Server system service on port 8787.
JupyterLab: Installs JupyterLab and commonly used packages (pip, numpy, pandas, matplotlib, ipython, and jupyter-ai) into a global Pixi environment (coder-jupyter). JupyterLab runs in the background on port 19999 using ~/workspace as the working directory.
Visual Studio Code (Desktop): Configures the Coder agent to enable desktop VS Code support. This surfaces convenient connection buttons in the Coder dashboard to launch a remote session directly into the workspace from your local Visual Studio Code application via the Coder extension.
Visual Studio Code (Web): Deploys the browser-based vscode-web service inside your workspace. It runs a web-accessible VS Code instance rooted at /home/coder/workspace.
DREAM Lab LLM API Key: Injects your DREAM Lab AI Gateway key into the workspace as $DREAMLAB_LLM_KEY. If set, the key is automatically used by OpenCode to authenticate requests to models provided through the gateway.

Connecting to Your Workspace

Once your workspace is created, you should be directed to the workspace dashboard. There are two main ways to interact with your new workspace: directly through the browser or by using software locally installed on your machine.

Workspace Dashboard with dotted red line around buttons for browser-based tools

Browser-based tools include:

JupyterLab, or RStudio: These are optional IDEs for data analysis using R or Python.
File Browser: for navigating your workspace’s file system and transferring files.
Terminal: a browser-based terminal that provides instant access to your workspace’s shell environment.

To use these, simply click the corresponding icon on your workspace’s dashboard page.

Connecting with Locally Installed Software

For a more integrated development experience, you can connect your local tools to your Coder workspace. This typically requires installing the Coder CLI and the appropriate extensions for your IDE.

Connecting via VS Code

You can develop in your Coder workspace remotely using the Visual Studio Code desktop client.

Ensure you have the Visual Studio Code desktop application installed.
Install the Coder extension from the VS Code Marketplace.
Once the extension is installed, open the Command Palette (Ctrl+Shift+P or Cmd+Shift+P on macOS) and type Coder: Login.
Enter your Coder deployment URL: https://coder.dreamlab.ucsb.edu .
After logging in, you can browse your workspaces in the Coder view on the sidebar, or use the Command Palette (Coder: Open Workspace) to connect to your workspace and start coding.

Connecting via Coder CLI

The Coder CLI can be used to authenticate and connect to your workspaces from the command line.

Installing the Coder CLI:

On Linux/macOS, the fastest way to install the CLI is using the install script:

curl -L https://coder.com/install.sh | sh

For Windows, download and run the installer program (coder_2.30.6_windows_amd64_installer.exe) from Coder’s GitHub Releases page. You may want to refer to the Coder CLI documentation.

Authenticate the CLI:

Once installed, log in to your Coder deployment using your terminal:

coder login https://coder.dreamlab.ucsb.edu

Follow the prompts to complete authentication via your web browser.

Connecting via SSH:

You can connect directly to your workspace using the Coder CLI:

coder ssh <workspace-name>