Install MCP Multi-Language Sandbox

Pick your client, copy the command, done.

Manual setup required. The maintainer's config contains paths only you know - edit the placeholders below before adding it to Claude Code.

Run this once before adding it to Claude Code.

git clone https://github.com/Pit-CL/mcp-multilang-sandbox.git
cd mcp-multilang-sandbox
npm install
npm run build

claude mcp add mcp-multilang-sandbox -- node /path/to/mcp-multilang-sandbox/dist/mcp/server.js

Replace any placeholder paths in the command with the real path on your machine.

Environment Variables

Set these before running MCP Multi-Language Sandbox.

VariableDescriptionRequired

LOG_LEVELSets the logging verbosity (e.g., info, debug)No

Once configured, MCP Multi-Language Sandbox gives your AI agent access to:

sandbox_executeExecute code in any supported language

languagecodesessionml

sandbox_sessionCreate/list/pause/resume/destroy sessions

actionnamelanguage

sandbox_installInstall packages with caching

sessionpackages

sandbox_file_opsRead/write/list/delete files in sessions

sessionoperationpathcontent

sandbox_inspectView system stats (pool, cache, sessions)

target

sandbox_securityView audit logs and security events

action

After setup, try these prompts with your AI agent:

→Execute a Python script to calculate the first 50 Fibonacci numbers.

→Create a new session named 'data-analysis' and install the pandas library.

→Write a TypeScript snippet to a file named 'test.ts' inside the 'data-analysis' session.

→Run a security audit to check the current status of the sandbox containers.

→Execute a Rust program that prints 'Hello from Rust' within a persistent session.

Prerequisites & system requirements

An MCP-compatible client (Claude Code, Cursor, Windsurf, Claude Desktop, or Codex)
Node.js 18+ with npm/npx installed

Save the docs, env vars, and workflow around MCP Multi-Language Sandbox in Conare so Claude Code, Codex, and Cursor remember it next time.