MCP Sandbox is an interactive Python code execution tool that enables users and language models to safely run Python code and install packages in isolated Docker containers. It provides a secure environment for code execution through containerization while offering package management capabilities and file generation support.
To set up the MCP Sandbox on your local machine:
# Clone the repository
git clone https://github.com/JohanLi233/python-mcp-sandbox.git
cd python-mcp-sandbox
uv venv
uv sync
# Start the server
uv run main.py
The server starts with the default SSE endpoint at http://localhost:8000/sse, which you can access via the MCP Inspector or any client that supports SSE connections.
# Create a new sandbox
create_sandbox() # Returns a sandbox_id for future operations
# List existing sandboxes
list_sandboxes() # Returns a list of available sandbox IDs
# Execute Python code in a sandbox
execute_python_code(sandbox_id="your_sandbox_id", code="print('Hello, World!')")
# Install a package in a sandbox
install_package_in_sandbox(sandbox_id="your_sandbox_id", package_name="pandas")
# Check installation status
check_package_installation_status(sandbox_id="your_sandbox_id", package_name="pandas")
# Execute terminal commands in a sandbox
execute_terminal_command(sandbox_id="your_sandbox_id", command="ls -la")
# Upload a file to a sandbox
upload_file_to_sandbox(
sandbox_id="your_sandbox_id",
local_file_path="/path/to/file.csv",
dest_path="/app/results" # Optional, defaults to /app/results
)
# List existing sandboxes
sandboxes = list_sandboxes()
# If no sandboxes exist, create a new one
if not sandboxes:
sandbox_response = create_sandbox()
sandbox_id = sandbox_response["sandbox_id"]
else:
sandbox_id = sandboxes[0]
# Install pandas and matplotlib
install_package_in_sandbox(sandbox_id=sandbox_id, package_name="pandas")
install_package_in_sandbox(sandbox_id=sandbox_id, package_name="matplotlib")
# Check installation status
pandas_status = check_package_installation_status(sandbox_id=sandbox_id, package_name="pandas")
matplotlib_status = check_package_installation_status(sandbox_id=sandbox_id, package_name="matplotlib")
code = """
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
# Create sample data
data = {'x': np.random.rand(100), 'y': np.random.rand(100)}
df = pd.DataFrame(data)
# Save data to CSV
df.to_csv('/app/results/data.csv', index=False)
# Create and save a plot
plt.figure(figsize=(10, 6))
plt.scatter(df['x'], df['y'])
plt.title('Random Scatter Plot')
plt.xlabel('X Axis')
plt.ylabel('Y Axis')
plt.savefig('/app/results/scatter_plot.png')
# Note: Do not use plt.show() as it won't work in the sandbox
print("Data and plot generated successfully!")
"""
result = execute_python_code(sandbox_id=sandbox_id, code=code)
# Result will contain stdout, stderr and any file_links to generated files
print("Output:", result["stdout"])
print("Errors:", result["stderr"])
print("Generated Files:", result["file_links"])
{
"mcpServers": {
"mcpSandbox": {
"command": "npx",
"args": ["-y", "supergateway", "--sse", "http://localhost:8000/sse"]
}
}
}
{
"mcpServers": {
"mcpSandbox": {
"command": "npx",
"args": ["-y", "supergateway", "--sse", "http://115.190.87.78/sse?api_key=<API_KEY>"]
}
}
}
When generating visualizations, don't use plt.show(). Instead use:
plt.savefig() to save figures to filesdf.to_csv() or df.to_excel() to save dataAll files should be saved to the /app/results directory to be accessible via links
Always check for existing sandboxes before creating new ones to efficiently manage resources
After installing packages, verify their installation status before using them in your code
To add this MCP server to Claude Code, run this command in your terminal:
claude mcp add-json "mcpSandbox" '{"command":"npx","args":["-y","supergateway","--sse","http://localhost:8000/sse"]}'See the official Claude Code MCP documentation for more details.
There are two ways to add an MCP server to Cursor. The most common way is to add the server globally in the ~/.cursor/mcp.json file so that it is available in all of your projects.
If you only need the server in a single project, you can add it to the project instead by creating or adding it to the .cursor/mcp.json file.
To add a global MCP server go to Cursor Settings > Tools & Integrations and click "New MCP Server".
When you click that button the ~/.cursor/mcp.json file will be opened and you can add your server like this:
{
"mcpServers": {
"mcpSandbox": {
"command": "npx",
"args": [
"-y",
"supergateway",
"--sse",
"http://localhost:8000/sse"
]
}
}
}To add an MCP server to a project you can create a new .cursor/mcp.json file or add it to the existing one. This will look exactly the same as the global MCP server example above.
Once the server is installed, you might need to head back to Settings > MCP and click the refresh button.
The Cursor agent will then be able to see the available tools the added MCP server has available and will call them when it needs to.
You can also explicitly ask the agent to use the tool by mentioning the tool name and describing what the function does.
To add this MCP server to Claude Desktop:
1. Find your configuration file:
~/Library/Application Support/Claude/claude_desktop_config.json%APPDATA%\Claude\claude_desktop_config.json~/.config/Claude/claude_desktop_config.json2. Add this to your configuration file:
{
"mcpServers": {
"mcpSandbox": {
"command": "npx",
"args": [
"-y",
"supergateway",
"--sse",
"http://localhost:8000/sse"
]
}
}
}3. Restart Claude Desktop for the changes to take effect