Announcing HopX MCP Server: Code Execution for AI Assistants

We're excited to announce the HopX MCP Server — a Model Context Protocol server that enables AI assistants like Claude, Cursor, and VS Code Copilot to execute code in secure, isolated cloud containers.

One command to install. Instant code execution superpowers.

bash

1	uvx hopx-mcp
2

What is MCP?

The Model Context Protocol (MCP) is an open standard that allows AI assistants to interact with external tools and services. Instead of just generating code, your AI can now run it.

With the HopX MCP Server, when you ask Claude to "analyze this data" or "test this function," it doesn't just write the code — it executes it in a secure sandbox and returns the actual results.

Why HopX MCP?

Before HopX MCP:

"Here's some Python code that should work. Copy it and run it yourself to see the output."

After HopX MCP:

"I ran the analysis. Here are the results: average is 42.5, median is 38, and I've generated a chart showing the distribution."

The difference is transformative. Your AI assistant becomes a true coding partner that can:

Execute code and see real output
Debug iteratively by running and fixing
Analyze data with pandas, numpy, matplotlib
Test solutions before presenting them
Work with files in isolated environments

Quick Start

1. Get Your API Key

2. Configure Your IDE

For Claude Desktop, add to your config file:

json

{
  "mcpServers": {
    "hopx-sandbox": {
      "command": "uvx",
      "args": ["hopx-mcp"],
      "env": {
        "HOPX_API_KEY": "your-api-key-here"
      }
    }
  }
}
 

Config locations:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

For Cursor, add to .cursor/mcp.json:

json

{
  "mcpServers": {
    "hopx-sandbox": {
      "command": "uvx",
      "args": ["hopx-mcp"],
      "env": {
        "HOPX_API_KEY": "your-api-key-here"
      }
    }
  }
}
 

For VS Code, add to .vscode/mcp.json:

json

{
  "mcpServers": {
    "hopx-sandbox": {
      "command": "uvx",
      "args": ["hopx-mcp"],
      "env": {
        "HOPX_API_KEY": "your-api-key-here"
      }
    }
  }
}
 

3. Start Using It

Restart your IDE and start asking your AI to run code!

Capabilities

Multi-Language Support

Execute code in multiple languages out of the box:

Language	Version	Pre-installed Packages
Python	3.11+	pandas, numpy, matplotlib, scipy, scikit-learn, requests
JavaScript	Node.js 20	Standard libraries
Bash	Latest	Unix utilities, git, curl, wget
Go	Latest	Compilation support

Available Tools

The MCP server provides these tools to your AI assistant:

Code Execution

execute_code_isolated() — One-shot execution (recommended)
execute_code() — Execute in existing sandbox
execute_code_background() — Long-running tasks
execute_code_async() — Webhook callbacks for 30+ min tasks

File Operations

file_read(), file_write(), file_list()
file_exists(), file_remove(), file_mkdir()

Shell Commands

run_command() — Execute shell commands
run_command_background() — Background processes

Sandbox Management

create_sandbox() — Create persistent sandbox
list_sandboxes() — List active sandboxes
delete_sandbox() — Terminate sandbox

Performance

We've optimized for the interactive AI use case:

Metric	Time
Sandbox creation	~200ms
Container startup	~100ms
Code execution	Near-instant

Your AI doesn't wait. Neither do you.

Security

Every code execution runs in complete isolation:

VM-level isolation — Each execution in its own micro-VM
Auto-cleanup — Containers destroyed after use (default: 10 min)
No cross-contamination — Sandboxes can't access each other
JWT authentication — Secure per-sandbox auth

Your local system is completely protected from executed code.

Example Use Cases

Data Analysis

Ask Claude: "Download the Titanic dataset and show me survival rates by passenger class"

Claude will:

Execute code to download the dataset
Run pandas analysis
Generate matplotlib visualizations
Return actual statistics and charts

Code Testing

Ask Cursor: "Write a function to validate email addresses and test it with edge cases"

Cursor will:

Write the validation function
Create test cases
Execute tests in a sandbox
Show real pass/fail results

Debugging

Ask your AI: "This code throws an error, fix it"

Your AI will:

Run the code to see the actual error
Analyze the traceback
Apply a fix
Run again to verify it works

Open Source

The HopX MCP Server is open source and available on GitHub:

github.com/hopx-ai/mcp

We welcome contributions, issues, and feedback from the community.

Get Started Today

Sign up at hopx.ai for your API key
Install with uvx hopx-mcp
Configure your IDE
Start building with AI-powered code execution

Transform your AI assistant from a code generator into a true coding partner.

Resources

Questions? Reach out at support@hopx.ai.

1	{
2	"mcpServers": {
3	"hopx-sandbox": {
4	"command": "uvx",
5	"args": ["hopx-mcp"],
6	"env": {
7	"HOPX_API_KEY": "your-api-key-here"
8	}
9	}
10	}
11	}
12