
ForeverVM
The sessionless code interpreter. Securely run AI-generated code in stateful sandboxes that run forever.
README
foreverVM
repo | version |
---|---|
cli | |
sdk |
foreverVM provides an API for running arbitrary, stateful Python code securely.
The core concepts in foreverVM are machines and instructions.
Machines represent a stateful Python process. You interact with a machine by running instructions (Python statements and expressions) on it, and receiving the results. A machine processes one instruction at a time.
Getting started
You will need an API token (if you need one, reach out to paul@jamsocket.com).
The easiest way to try out foreverVM is using the CLI. First, you will need to log in:
npx forevervm login
Once logged in, you can open a REPL interface with a new machine:
npx forevervm repl
When foreverVM starts your machine, it gives it an ID that you can later use to reconnect to it. You can reconnect to a machine like this:
npx forevervm repl [machine_name]
You can list your machines (in reverse order of creation) like this:
npx forevervm machine list
You don't need to terminate machines -- foreverVM will automatically swap them from memory to disk when they are idle, and then automatically swap them back when needed. This is what allows foreverVM to run repls “forever”.
Using the API
import { ForeverVM } from '@forevervm/sdk'
const token = process.env.FOREVERVM_TOKEN
if (!token) {
throw new Error('FOREVERVM_TOKEN is not set')
}
// Initialize foreverVM
const fvm = new ForeverVM({ token })
// Connect to a new machine.
const repl = fvm.repl()
// Execute some code
let execResult = repl.exec('4 + 4')
// Get the result
console.log('result:', await execResult.result)
// We can also print stdout and stderr
execResult = repl.exec('for i in range(10):\n print(i)')
for await (const output of execResult.output) {
console.log(output.stream, output.data)
}
process.exit(0)
Working with Tags
You can create machines with tags and filter machines by tags:
import { ForeverVM } from '@forevervm/sdk'
const fvm = new ForeverVM({ token: process.env.FOREVERVM_TOKEN })
// Create a machine with tags
const machineResponse = await fvm.createMachine({
tags: {
env: 'production',
owner: 'user123',
project: 'demo'
}
})
// List machines filtered by tags
const productionMachines = await fvm.listMachines({
tags: { env: 'production' }
})
Memory Limits
You can create machines with memory limits by specifying the memory size in megabytes:
// Create a machine with 512MB memory limit
const machineResponse = await fvm.createMachine({
memory_mb: 512,
})
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.