This skill uses the E2B sandbox CLI to let an AI agent create and manage isolated Linux environments for running code, testing packages, executing commands, and working with binary files. It is built for developers who need a safe, disposable workspace for experimental or untrusted code. The sandbox supports file uploads and downloads, browser automation for UI validation, and multi-agent setups where each agent tracks its own sandbox independently.
Follow the documented setup, then try a first task.
Installs: Complete Agent Sandbox Skill repository setup · Claude Code, Gemini CLI, Codex CLI
Run from: Project root of the cloned repository
Clone the repository if it has not already been cloned. The source does not provide a clone command.
From the project root, create the environment file from the supplied sample.
cp .env.sample .env
Edit `.env` and replace `sbx_...` with your E2B API key obtained from the E2B Dashboard.
To use Claude Code, open a terminal in the configured project and launch it.
claude
Alternatively, to use Gemini CLI, open a terminal in the configured project and launch it.
gemini
Alternatively, to use Codex CLI, open a terminal in the configured project and launch it.
codex
Uses Agent Sandboxes
Use the Agent Sandboxes skill to run the following Python snippet in an isolated E2B sandbox and report its stdout and any errors: `<python-code>`. Validate that the configured E2B API key is available before starting, and do not create files on my local system.
Claude
Codex
Gemini CLI