SMART-E2B

SMART-E2B is an MCP server that integrates the E2B platform to enable secure code execution inside isolated cloud sandboxes for Claude AI Desktop. Designed for software developers, data scientists, and engineers using AI assistants, the server provides a controlled runtime environment where untrusted or dynamic scripts can run without risking local host machines. Through this connection, Claude can directly execute JavaScript and Python programs, inspect execution outputs, and handle errors or timeouts automatically. SMART-E2B also includes integrated file system operations, allowing assistants to upload files into the sandbox, inspect directory trees, and read remote file contents. Furthermore, the server features intelligent sandbox session reuse to optimize execution latency and preserve environment state between related commands. By bridging conversational AI with scalable cloud execution, SMART-E2B allows users to perform data analysis, test code snippets, generate dynamic charts, and prototype scripts directly within desktop conversations while maintaining strict security boundaries.

Category: Cloud & Infrastructure

Tags: cloud computing, code-execution, e2b, sandbox

Visit SMART-E2B

How to install and configure SMART-E2B

To set up SMART-E2B with Claude AI Desktop, ensure Node.js (v16 or higher) is installed and obtain an API key from e2b.dev. 1. Install SMART-E2B automatically via Smithery: bash npx -y @smithery/cli install @Leghis/smart-e2b --client claude Alternatively, install it globally using npm: bash npm install -g smart-e2b 2. Open Claude AI Desktop, navigate to Settings > Developer, and select Edit Config to open claude_desktop_config.json. 3. Add the server under mcpServers: json { "mcpServers": { "smart-e2b": { "command": "npx", "args": ["-y", "smart-e2b"], "env": { "E2B_API_KEY": "<YOUR_E2B_API_KEY>" } } } } 4. Save the file and restart Claude AI Desktop.

What you can do with SMART-E2B

• Running Python data analysis scripts using libraries like NumPy and Matplotlib to calculate dataset statistics and generate visualizations. - Testing isolated JavaScript algorithms and helper functions inside a safe cloud runtime without installing local dependencies. - Uploading input files to remote cloud sandboxes to process structured datasets and extract transformed results. - Inspecting sandbox directory contents and reading generated output files directly from the Claude conversation interface. - Prototyping multi-step automation scripts across persistent cloud sandbox sessions that preserve working state between commands.

Key facts

• Freemium
• https://github.com/Leghis/smart-e2b
• Cloud & Infrastructure, Developer Tools & Code Intelligence
• cloud computing, code-execution, e2b, sandbox

Part of MCP Servers

How do I install SMART-E2B?

You can install SMART-E2B automatically for Claude Desktop using the Smithery CLI with the command npx -y @smithery/cli install @Leghis/smart-e2b --client claude. Alternatively, install it globally using npm install -g smart-e2b or clone the GitHub repository and configure the npx execution block in your Claude Desktop configuration file.

What can SMART-E2B do?

SMART-E2B exposes tools for running JavaScript and Python code inside cloud sandboxes managed by E2B. It includes executeJavaScript, executePython, uploadFile, listFiles, and readFile. The server also handles timeouts, captures standard outputs and errors, and intelligently reuses sandbox sessions across conversational turns to preserve execution context.

Which MCP clients work with SMART-E2B?

SMART-E2B is specifically configured and documented for Claude AI Desktop. Because it communicates via standard Model Context Protocol conventions over stdio, it can also be configured with other MCP-compliant hosts that allow setting custom command arguments and environment variables.

Is SMART-E2B open source?

Yes, SMART-E2B is open source and distributed under the MIT license. You can inspect its source code, build it locally, or contribute improvements directly on its official GitHub repository at https://github.com/Leghis/smart-e2b. Note that running sandboxes requires an active E2B account and API key.

What prerequisites are required for SMART-E2B?

Running SMART-E2B requires Node.js version 16 or higher, Claude AI Desktop or a compatible MCP client, and an active E2B API key from e2b.dev. The API key must be defined in your client configuration file under the E2B_API_KEY environment variable.