ForeverVM Python REPL MCP Server

Data Science ToolsJavaScript

Execute Python code in a persistent REPL environment

Available Tools

create-python-repl

Create a new Python REPL instance that persists throughout the conversation

run-python-in-repl

Execute Python code in a specific REPL instance

codereplId

ForeverVM Python REPL provides a persistent Python execution environment that allows Claude to run and maintain state across multiple Python code executions. This MCP server creates isolated Python REPL instances that persist throughout your conversation, enabling complex data analysis, visualization, and computation tasks without losing variable context between runs.

Overview

ForeverVM Python REPL provides a persistent Python execution environment for Claude, allowing you to run Python code and maintain state throughout your conversation. This is particularly useful for data analysis, scientific computing, and other tasks that require multiple code executions that build on previous results.

Installation

For Claude Desktop

The simplest way to install ForeverVM Python REPL is by running:

npx forevervm-mcp install --claude

This command automatically configures the MCP server for Claude Desktop.

For Other MCP Clients

To use ForeverVM Python REPL with other MCP clients, add the following configuration to your client settings:

{
  "mcpServers": {
    "forevervm-python": {
      "command": "npx",
      "args": ["forevervm-mcp", "run"]
    }
  }
}

For Development (Local Installation)

If you're developing or modifying the MCP server, you can run it locally:

Clone the repository
Navigate to the mcp-server directory
Configure your MCP client with:

{
  "mcpServers": {
    "forevervm-python": {
      "command": "npm",
      "args": ["--prefix", "PATH_TO_REPOSITORY/javascript/mcp-server", "run", "start", "run"]
    }
  }
}

Replace PATH_TO_REPOSITORY with the actual path to your local repository.

Usage

Once installed, you can use the Python REPL in your conversations with Claude. The MCP server provides two main tools:

create-python-repl: Creates a new Python REPL instance
run-python-in-repl: Executes Python code in a specific REPL instance

Claude will automatically handle the creation and management of REPL instances, so you can simply ask it to run Python code. For example:

"Can you create a simple plot using matplotlib in Python?"

Claude will:

Create a new REPL if needed
Run the appropriate Python code
Return the results to you

The REPL maintains state throughout your conversation, so variables, functions, and imported libraries persist between code executions.

For more detailed documentation, visit the ForeverVM docs.