MCP QuickJS Runner
Provides secure execution of arbitrary JavaScript code within a sandboxed QuickJS WASM environment, allowing language models or other MCP clients to safely run JavaScript code snippets without compromising the host system.
README
MCP QuickJS Runner
A server implementing the Model Context Protocol (MCP) that provides a tool to securely execute arbitrary JavaScript code within a QuickJS engine compiled to WebAssembly (WASM) and run using Node.js's built-in WASI implementation.
Description
This server acts as an MCP tool provider. It exposes a single tool, run_javascript_code, which takes a string of JavaScript code as input. The code is then executed inside a sandboxed QuickJS WASM environment. The server captures the standard output (stdout) and standard error (stderr) streams from the execution and returns them, along with any execution errors, back to the MCP client.
This allows language models or other MCP clients to safely execute potentially untrusted JavaScript code snippets without compromising the host system.
Features
- Secure Execution: Runs JavaScript in a WASM sandbox using QuickJS and Node.js WASI.
- Standard I/O Capture: Captures
stdoutandstderrfrom the executed JavaScript code. - Error Reporting: Reports runtime errors from QuickJS and non-zero exit codes.
- MCP Integration: Exposes functionality as a standard MCP tool over
stdio. - Built with TypeScript: Provides type safety during development.
How it Works
- WASM Module: Uses a pre-compiled QuickJS engine (
qjs-wasi.wasm) targeting the WebAssembly System Interface (WASI). - Node.js WASI: Leverages the
node:wasimodule in Node.js to instantiate and run the WASM module. - Stdio Redirection (Temporary Files): To capture
stdoutandstderrfrom the WASM environment, the server currently relies on the standard approach compatible withnode:wasi:- A temporary directory is created on the host filesystem using
node:fs/promisesandnode:os. - Temporary files for
stdoutandstderrare opened within this directory. - The real OS file descriptors for these files are passed to the
WASIinstance during initialization (stdout: fd,stderr: fd). - The QuickJS WASM module writes its output to these descriptors, which gets routed by WASI to the temporary files.
- After execution finishes, the server closes the file handles and reads the content of the temporary files.
- The temporary directory and files are cleaned up.
- (Note: Attempts to use in-memory pipes or virtual filesystems like
memfswere unsuccessful becausenode:wasicurrently requires real OS file descriptors for stdio.)
- A temporary directory is created on the host filesystem using
- MCP Communication: The server uses
@modelcontextprotocol/sdkto listen for MCP requests viastdioand respond with the execution results formatted according to the protocol.
Prerequisites
- Node.js (v23.x or later recommended, check
node:wasicompatibility for your specific version) - npm or yarn
- The QuickJS WASM file (
qjs-wasi.wasm) must be present in the same directory as the compiled server script (e.g.,./dist/qjs-wasi.wasmrelative to./dist/server.js). You may need to obtain or compile this separately.
Installation
- Clone the repository (if applicable).
- Install dependencies:
npm install
Usage
node server.ts
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.