e2b-desktop
Control E2B Desktop sandboxes (virtual Linux desktops) for computer-use agents. Use when you need to create/manage sandboxed desktop environments, take screenshots, perform mouse/keyboard actions, run commands, stream VNC output, or build computer-use agent loops with E2B Desktop SDK.
Why use this skill?
Use the e2b-desktop skill to control headless Linux environments with OpenClaw. Automate UI testing, screen capture, and desktop tasks easily.
Install via CLI (Recommended)
clawhub install openclaw/skills/skills/eyhn/e2b-desktopWhat This Skill Does
The e2b-desktop skill provides a powerful interface for controlling headless Ubuntu Linux sandboxes (equipped with XFCE) directly through OpenClaw. It serves as a bridge for computer-use agents, allowing them to interact with a graphical desktop environment as if they were a human user. By wrapping the robust E2B Python SDK in a set of bash-accessible scripts, this skill enables agents to perform complex tasks such as launching applications, clicking UI elements, dragging objects, typing, and executing shell commands within an isolated, reproducible environment. It also supports real-time monitoring through VNC stream capabilities and state persistence via local session management.
Installation
To integrate this skill into your OpenClaw ecosystem, execute the following command in your terminal:
clawhub install openclaw/skills/skills/eyhn/e2b-desktop
Ensure you have the Python dependencies installed within your agent's environment:
pip install e2b-desktop
You must also set your environment variable to authenticate with the E2B platform:
export E2B_API_KEY=e2b_***
Use Cases
- Automated UI Testing: Run a suite of tests by launching browsers or desktop apps within the sandbox and validating UI states via screenshots.
- Complex Data Extraction: If a website or application lacks an API, use this skill to open the app, navigate the interface, and copy data.
- Sandboxed Development: Perform exploratory coding or build tasks in a clean Linux environment without touching your local machine's files.
- Agentic Workflows: Build autonomous loops where the agent takes a screenshot, analyzes the UI, and performs clicks/typing to achieve a high-level user goal.
Example Prompts
- "Open the browser in the E2B desktop, navigate to github.com, and take a screenshot of the landing page."
- "Start a new desktop sandbox at 1920x1080, launch a terminal, and run 'ls -la' to see the current directory."
- "Help me click the 'Submit' button on the form shown in the latest screenshot."
Tips & Limitations
- State Management: The skill automatically tracks sandbox IDs in
~/.e2b_state. Always usekill_sandbox.shwhen finished to prevent incurring costs for idle sandboxes. - Latency: Because this skill operates via a remote cloud sandbox, mouse movements and screenshot capture will have slight latency. Implement adequate
wait.shcommands after launching apps to ensure the GUI is fully loaded. - Persistence: While sandboxes survive script exit, they are ephemeral in nature. Do not rely on them for long-term file storage without external syncing.
Metadata
Not sure this is the right skill?
Describe what you want to build — we'll match you to the best skill from 16,000+ options.
Find the right skillPaste this into your clawhub.json to enable this plugin.
{
"plugins": {
"official-eyhn-e2b-desktop": {
"enabled": true,
"auto_update": true
}
}
}Tags(AI)
Flags: network-access, file-write, file-read, external-api, code-execution