A WebAssembly runtime and CLI for running apps and AI agent code in fast, lightweight, sandboxed environments.
taybenlor/runno
A sandboxed WebAssembly runtime for running code in multiple languages in the browser, on Node, or through an MCP server.
- Stars
- 775
- Forks
- 48
- Open issues
- 38
- Last commit
- 7 days ago
What is it?
Runno is a set of JavaScript packages for executing code in various programming languages inside a sandbox. It uses WASI, the WebAssembly System Interface, to emulate a unix-like environment with a virtual file system, so programs run entirely within JavaScript without touching the host's real files or network.
The @runno/runtime package provides web components that let a page run code examples directly in the browser, with no server needed. The @runno/sandbox package runs code in Node and other JavaScript runtimes, and @runno/wasi runs WASI binaries in either environment. @runno/mcp exposes the sandbox as an MCP server so AI assistants can execute code.
Supported runtimes include Python, Ruby, QuickJS, SQLite, C, C++, and PHP CGI. Because WASI does not provide a full Linux environment, Runno is best suited to small programs that read from stdin and write to stdout, and installing external packages is difficult.
Key features
- 01Runs code in multiple languages including Python, Ruby, JavaScript, C, C++, and SQLite.
- 02Provides web components for embedding runnable code examples in browser pages.
- 03Runs code in Node and other JavaScript runtimes via a sandbox API.
- 04Includes an MCP server so AI assistants can execute code.
- 05Uses a virtual file system with no real file or network access.
Use cases
- Making programming course examples runnable in the browser without installing tools.
- Testing small code snippets safely inside a sandbox on a user's machine.
- Giving an AI assistant a sandboxed place to execute code via MCP.
Quick start
- 1
Install the sandbox package
npm install @runno/sandbox - 2
Run Ruby code from your script
- 3
Add the MCP server to your MCP client config
Things to know
- Supported languages are limited to those already compiled to WebAssembly with WASI.
- WASI provides no full Linux environment, and installing packages or modules inside Runno is difficult.
- Best suited to small programs that use STDIN and STDOUT.
In collections
Topics
FAQ
- What is runno?
- A sandboxed WebAssembly runtime for running code in multiple languages in the browser, on Node, or through an MCP server. Runno is a set of JavaScript packages for executing code in various programming languages inside a sandbox. It uses WASI, the WebAssembly System Interface, to emulate a unix-like environment with a virtual file system, so programs run entirely within JavaScript without touching the host's real files or network.
- Is runno free to use? What license does it have?
- runno is open source under the MIT license. Check the license terms before using it in commercial products.
- Is runno still maintained?
- Actively maintained. The last commit to runno was on Oct 31, 2026.
- What can I use runno for?
- Making programming course examples runnable in the browser without installing tools. Testing small code snippets safely inside a sandbox on a user's machine. Giving an AI assistant a sandboxed place to execute code via MCP.
- How do I install runno?
- Install the sandbox package: npm install @runno/sandbox Run Ruby code from your script Add the MCP server to your MCP client config
- Where can I get runno?
- The source code and installation instructions are on GitHub: https://github.com/taybenlor/runno
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A standalone WebAssembly runtime built on the Cranelift code generator, with WASI support and embeddings for multiple languages.
A fast WebAssembly interpreter in C, small enough for microcontrollers and embeddable as a library in many languages.
A library and hosted service that gives AI agents isolated Linux-like sandboxes with files, processes, shell, and networking, built on WebAssembly and V8.
A runtime that executes untrusted code for AI agents inside isolated WebAssembly sandboxes, with CPU, memory and timeout limits per task.
An all-in-one Docker sandbox giving AI agents a shared browser, shell, file system, VSCode Server, Jupyter and MCP services.