Agent Runtime Infrastructure · Preview

E2B Sandbox Code Execution Pack

E2B capability pack for isolated sandboxes where AI agents can safely execute code, process data, run commands, and manage files.

Best fitDevelopers who need isolated code execution for AI agents, data processing, generated files, or command-running workflows.

Check whether this project matches your task before installing it.

What it can doSandbox execution preflight, filesystem and network boundary review, command acceptance checks, timeout budget, and recovery guidance

Review the portable capability path.

Before continuingVerify in a sandbox

Do not treat a preview pack as a proven local install.

GitHub snapshot12k stars

907 forks · 50 contributors

Official first step Read manual preview Source repository

Preview status · 2026-05-20

What is E2B Sandbox Code Execution Pack?

E2B provides secure cloud sandboxes where AI agents and apps can execute code, run commands, process data, and manage files.
Best fit: Developers who need isolated code execution for AI agents, data processing, generated files, or command-running workflows.
Not for: Not for unbounded shell access or workflows that cannot isolate credentials, files, network, and runtime cost.
Capability added to an AI workflow: Sandbox execution preflight, filesystem and network boundary review, command acceptance checks, timeout budget, and recovery guidance
First safe verification step: Run one disposable sandbox task with a timeout, tiny input, no secrets, and a clear cleanup path.
Verification state: source, Quick Start, and sandbox install checks are recorded as passed.
Top risk: The main risk is letting an agent execute code without boundaries for files, credentials, network, timeout, or cost.
Evidence base: https://github.com/e2b-dev/E2B, https://github.com/e2b-dev/E2B#readme, Human Manual, Pitfall Log

Quick decision

Use this section to decide whether the project is worth a deeper read.

Best forDevelopers who need isolated code execution for AI agents, data processing, generated files, or command-running workflows.

Match the project to your task before installing it.

CapabilitySandbox execution preflight, filesystem and network boundary review, command acceptance checks, timeout budget, and recovery guidance

E2B capability pack for isolated sandboxes where AI agents can safely execute code, process data, run commands, and manage files.

Repositorye2b-dev/E2B

12k stars · 907 forks

What it can do

Translate the upstream project into concrete capabilities the user can judge before installing.

Overview

Related topics: Sandbox Usage, JavaScript/TypeScript SDK, Python SDK

Sources: [README.md:1]()

Installation

Related topics: JavaScript/TypeScript SDK, Python SDK, CLI Tool

Source: https://github.com/e2b-dev/E2B / Human Manual

Sandbox Usage

Related topics: Sandbox Management, Filesystem and Git Operations, Network Configuration

Sources: [README.md:1-20](https://github.com/e2b-dev/E2B/blob/main/README.md)

JavaScript/TypeScript SDK

Related topics: Python SDK, CLI Tool, Sandbox Usage

Sources: [packages/js-sdk/tsup.config.js]()

Python SDK

Related topics: JavaScript/TypeScript SDK, CLI Tool, Sandbox Usage

Sources: [e2b/api/metadata.py](https://github.com/e2b-dev/E2B/blob/main/packages/python-sdk/e2b/api/metadata.py)

Sources: https://github.com/e2b-dev/E2B, Human Manual, Project Pack evidence, and downstream validation signals.

Community Discussion Evidence

Project-level external discussion stays visible on the detail page, not only inside the manual.

Stars12k stars

Forks907 forks

Contributors50 contributors

Licenseunknown

Community Discussion Evidence

12 source-linked items

Review these external discussions before using E2B Sandbox Code Execution Pack with real data or production workflows. They are review inputs, not standalone proof that the project is production-ready.

01
Paused sandbox is not persisting the file changes / addition after secon
github / github_issue
02
(feature request) Run-scoped messaging for multi-sandbox agent workflows
github / github_issue
03
process was not killed when auto paused
github / github_issue
04
Docker Build Secrets Support
github / github_issue
05
When using autoPause in sandbox creation, connect overrides this when re
github / github_issue
06
Closed Port Error
github / github_issue
07
AuthenticationError: Unauthorized, please check your credentials. - Inva
github / github_issue
08
Template Build Fails with "syncing took too long" Error
github / github_issue
09
build status polling timed out
github / github_issue
10
[Bug]: Incorrect info about webhooks in docs and dashboard
github / github_issue
11
How to use the file that I uploaded when I create template
github / github_issue
12
[Bug]: Sandbox create fails from template, but succeeds without template
github / github_issue

How to start

Only source-backed commands are shown here. Verify them in an isolated environment first.

Try the prompt first

Test the workflow without installing the upstream project.

preview

Read the Human Manual

Understand inputs, outputs, limits, and failure modes.

manual

Take context to your AI host

Use the compiled assets in your preferred AI environment.

context

Run sandbox verification

Confirm install commands and rollback before using a primary environment.

verify

npm i e2b

Official start command · https://github.com/e2b-dev/E2B#readme · verified: yes

Human Manual

The English page must expose the real manual, not a short placeholder.

8+ sections · Human Manual