Runno Code Sandbox
@taybenlor
Runno Code Sandbox について
The @runno/mcp server provides a run_code tool that AI assistants can use to execute code snippets in various programming languages. It leverages the @runno/sandbox package, which uses WebAssembly to create isolated environments for code execution. The result of the code executio
基本情報
設定
以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。
{
"mcpServers": {
"runno": {
"command": "/usr/local/bin/npx",
"args": [
"@runno/mcp"
]
}
}
}ツール
ツールは検出されませんでした
ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。
概要
What is Runno Code Sandbox?
Runno Code Sandbox helps you make runnable code examples that can be embedded in web pages. It uses a WASI (Web Assembly System Interface) runtime to emulate a file system and operating system inside the browser, allowing programs to run without modification. It is primarily designed for programming education, enabling learners to run code without installing complex tools and creators to embed interactive examples without a server.
How to use Runno Code Sandbox?
Install the @runno/runtime npm package, import it in your entrypoint file, and set the HTTP headers Cross-Origin-Opener-Policy: same-origin and Cross-Origin-Embedder-Policy: require-corp to create a cross-origin isolated context. Then use the <runno-run> element with a supported runtime (e.g., runtime="python") and your code as its content. For running WASI binaries directly, use <runno-wasi>.
Key features of Runno Code Sandbox
- Runs code in the browser using WebAssembly and WASI.
- Provides a sandboxed, in-memory file system.
- Supports multiple runtimes: Python, Ruby, QuickJS, SQLite, Clang, Clang++, PHP-CGI.
- Offers both web components (
@runno/runtime) and a headless WASI runtime (@runno/wasi). - Code executes in a Web Worker with terminal emulation via XTerm.js.
- No server-side execution required—all processing happens on the client.
Use cases of Runno Code Sandbox
- Creating interactive programming tutorials or documentation with live code examples.
- Allowing new programmers to run code without installing compilers or interpreters.
- Testing simple programs for correctness inside a client-side sandbox.
- Embedding runnable code snippets in blog posts or educational websites.
- Compiling and running C/C++ code directly in the browser using Clang runtimes.
FAQ from Runno Code Sandbox
How does Runno run code in the browser?
Runno fetches the WebAssembly binary for the chosen language from WAPM (Web Assembly Package Manager), then runs it in a Web Worker inside a sandbox with an in-memory file system. STDIN/STDOUT are connected to a terminal emulator.
What programming languages are supported?
Supported runtimes include Python, Ruby, QuickJS (JavaScript), SQLite, Clang (C), Clang++ (C++), and PHP-CGI. Availability depends on what has been compiled to WASI and uploaded to WAPM.
What security measures does Runno have?
Two sandboxing layers are used: the WebAssembly layer controls system resource access, and embedding within a cross-origin iframe isolates the code from the hosting page’s cookies and API calls. However, resource hogging (e.g., infinite loops) remains possible.
What are the limitations of Runno?
Not all programming languages are available; only those compiled to WASI. There is no full Linux environment, so system-level interfaces often won’t work. Importing packages or modules is difficult; code must be provided inline. It is best suited for small examples using STDIN/STDOUT.
What HTTP headers are required for Runno to work?
You must set Cross-Origin-Opener-Policy: same-origin and Cross-Origin-Embedder-Policy: require-corp to enable SharedArrayBuffer, which is needed for STDIN support.
「推論」の他のコンテンツ
ArduPilot MCP Server Sandbox
hfujikawa77ArduPilotドローンをAIエージェントから操作するMCPサーバーです。
n8n Workflow Builder MCP
ifmelateMCP server that allow LLM in agent mode builds n8n workflows for you
Sandbox Mcp
pottekkatA Model Context Protocol (MCP) server that enables LLMs to run ANY code safely in isolated Docker containers.
Code Reasoning MCP Server
mettamattA code reasoning MCP server, a fork of sequential-thinking
AWorld: Advancing Agentic AI
inclusionAISearch, understand, reproduce, and improve an idea with ease
コメント