
Code Runner
- 242 repo stars
- Updated February 5, 2026
- formulahendry/mcp-server-code-runner
io.github.formulahendry/code-runner is a MCP server that runs multi-language code on demand for agent-driven development loops.
About
Code Runner is an MCP server from formulahendry that lets AI coding agents execute code in various programming languages through the Model Context Protocol. developers who ship with Claude Code, Cursor, Codex, or Windsurf install it when they want the agent to prove a snippet works, reproduce a bug in isolation, or compare outputs without manual copy-paste into a local REPL. The package publishes to npm as mcp-server-code-runner and also offers a Docker image for teams that prefer OCI-based MCP hosting. Transport is stdio, which matches the default pattern in most desktop agent configs. It is a narrow integration, not a full test runner or CI substitute: you still own sandboxing, secrets, and production safety. Used well, it shortens the loop between suggestion and verified behavior during backend experiments, CLI utilities, and agent-built prototypes.
- Runs code across multiple programming languages from MCP tool calls
- Ships as npm package mcp-server-code-runner with stdio MCP transport
- Optional OCI image docker.io/formulahendry/mcp-server-code-runner for containerized installs
- Server schema version 0.1.9 with GitHub source at formulahendry/mcp-server-code-runner
- Fits validate-then-build loops when agents need empirical output, not guessed results
Code Runner by the numbers
- Data as of Jul 7, 2026 (Skillselion catalog sync)
claude mcp add mcp-server-code-runner -- npx -y mcp-server-code-runnerAdd your badge
Show developers this MCP server is listed on Skillselion. Paste this into your README.
| repo stars | ★ 242 |
|---|---|
| Package | mcp-server-code-runner |
| Transport | STDIO |
| Auth | None |
| Last updated | February 5, 2026 |
| Repository | formulahendry/mcp-server-code-runner ↗ |
What it does
Let your coding agent execute snippets and scripts in many languages without leaving the chat thread.
Who is it for?
Best when you want quick in-chat execution while prototyping APIs, scripts, and small backend logic.
Skip if: Skip if you need hardened multi-tenant sandboxes, long-running jobs, or production-grade CI instead of ad hoc runs.
What you get
After you register the server, your agent can run snippets via MCP and base next steps on real stdout, stderr, and exit behavior.
- MCP tools that execute agent-supplied code and return run output
- Faster verify-and-iterate loops during build and ship debugging
- Optional containerized MCP deployment via the OCI package
By the numbers
- Published server version 0.1.9 in server metadata
- Two distribution paths: npm mcp-server-code-runner and OCI docker.io/formulahendry/mcp-server-code-runner:0.1.8
- Stdio MCP transport on both packages
README.md
Code Runner MCP Server
MCP Server for running code snippet and show the result.
It supports running multiple programming languages: JavaScript, PHP, Python, Perl, Perl 6, Ruby, Go, Lua, Groovy, PowerShell, BAT/CMD, BASH/SH, F# Script, C# Script, VBScript, TypeScript, CoffeeScript, Scala, Swift, Julia, Crystal, OCaml Script, R, AppleScript, Elixir, Clojure, Racket, Scheme, AutoHotkey, AutoIt, Kotlin Script, Dart, Haskell, Ni, Lisp, Kit, V, SCSS, Sass. Full list could be seen here in constants.ts.
Setup
npx for VS Code
Install the Code Runner MCP server in VS Code using below buttons:
Alternatively, you can add configuration in settings.json:
{
"mcp": {
"inputs": [],
"servers": {
"mcp-server-code-runner": {
"command": "npx",
"args": [
"-y",
"mcp-server-code-runner@latest"
],
}
}
}
}
npx for Claude Desktop
Configuration in claude_desktop_config.json:
{
"mcpServers": {
"mcp-server-code-runner": {
"command": "npx",
"args": [
"-y",
"mcp-server-code-runner@latest"
],
}
}
}
Docker
Use VS Code as example. Install the Code Runner MCP server in VS Code using below buttons:
Alternatively, you can add configuration in settings.json:
{
"mcp": {
"inputs": [],
"servers": {
"mcp-server-code-runner": {
"command": "docker",
"args": [
"run",
"--rm",
"-i",
"formulahendry/mcp-server-code-runner"
]
}
}
}
}
npx issue on Widnows
On Windows, MCP servers may fail to connect with npx.
You could try below two workarounds:
use bunx
- Install Bun.
- In configuration, change
npxwithbunx.
use cmd
Below is VS Code configuration in settings.json:
{
"mcp": {
"inputs": [],
"servers": {
"mcp-server-code-runner": {
"command": "cmd",
"args": [
"/c",
"npx",
"-y",
"mcp-server-code-runner@latest"
],
}
}
}
}
Run with Streamable HTTP Transport
npm install -g mcp-server-code-runner@latest
mcp-server-code-runner --transport http
Usage
Before using Code Runner MCP Server, please make sure interpreter or compiler of the programming language you want to run is set in PATH environment variable.
Try below prompts in the application which has configured Code Runner MCP Server:
Run the JavaScript Code: console.log(5+6)Where is temporary folder in my OS? Use run-code toolHow many CPUs do I have in my machine? Use run-code tool


Build your own MCP Server
Want to build your own MCP Server? Try Yeoman Generator for MCP Server to create your MCP Server project!
Recommended MCP Servers
How it compares
MCP code-execution bridge, not a hosted IDE or full continuous-integration platform.
FAQ
Who is code-runner for?
Developers using MCP-capable agents who want fast, language-agnostic snippet execution during build and debug sessions.
When should I use code-runner?
Use it when an agent-generated script, algorithm, or API example needs to be executed and interpreted before you merge or ship changes.
How do I add code-runner to my agent?
Install mcp-server-code-runner from npm or pull the formulahendry Docker image, add a stdio MCP server entry in your agent config, then restart the client so tools appear.