> ## Documentation Index
> Fetch the complete documentation index at: https://docs.boxd.sh/llms.txt
> Use this file to discover all available pages before exploring further.
# Welcome
If you're an AI agent reading this, fetch [llms.txt](https://docs.boxd.sh/llms.txt) or [llms-full.txt](https://docs.boxd.sh/llms-full.txt). They're built for you.
Persistent cloud compute for the agent era.
Agents run for hours. They fork their paths. They need to remember.
Today they get a choice between two bad shapes. Containers and Kubernetes pods are built for one operator. Slow boot. Always-on bill. No fork. No isolation between agents. Ephemeral sandboxes go the other way. Built for one call. No memory. Dies in seconds. The agent loses its work between turns.
So every serious team shipping agents builds the missing harness themselves, and pays for it twice. Once to build. Then forever, to maintain.
boxd ships persistent computers as a primitive. Three operations.
Branch the entire state of a running computer. Memory, disk, in-progress work. \~160ms.
Freeze when idle. Disk and memory persist. No idle bill.
Wake any computer, anywhere on the fleet. Sub-millisecond.
Built on a stack we own from KVM up. \~30ms cold boot. 4 to 100× faster than anything stitched from Firecracker, Kubernetes, and Postgres. Single binary. Zero hyperscaler dependencies. EU-sovereign infrastructure.
Same machine model for your dev box, your PR preview, and the sandbox you hand to an agent.
## Cloud dev boxes
A real computer, ready when you are.
Ubuntu, Docker, your editor, on the internet in under a second.
Warm copies of your app, on demand, in \~160ms.
A real `https://name.boxd.sh` the moment your server starts.
Cursor, VS Code, JetBrains, Zed. SSH straight in.
## Preview environments
A URL per branch. A URL per PR. A URL per anything.
Every PR gets its own boxd. Destroyed on merge.
Push to main, see it live in seconds. One GitHub webhook.
## Agents on boxd
The cloud computer agents actually want. Full root, persistent disk, a public HTTPS URL, and fork to recover.
Give Claude its own VM. Fork before risky moves. Destroy after.
Label an issue `boxd-fix`. Get a PR with a live preview URL.
Pre-installed. Persistent auth. Reads your `CLAUDE.md` by default.
Pre-installed. Same auth model, same primitives, `AGENTS.md` aware.
Pre-installed. Open-source agent harness on a real Linux box.
## Organizations
One bill, one fleet, one team. Pool machines under an org and share any of them with a single command — with a clean credential handoff every time.
Shared billing and machines for a team. Personal, org-billed, or shared.
Open a machine to your team. Personal logins never leak — guaranteed.
## Integrations
Connect Linear, Slack, and GitHub once. Every personal machine can use them, and the coding agents on board discover them automatically.
Clone and push private repos from any machine.
Linear's GraphQL API and an optional MCP server.
The Slack Web API, plus an optional MCP server.
## Get started
From zero to a live machine in under a minute.
The VM model, fork semantics, HTTPS proxying, suspend and resume.