fleet

io.github.lion-zhangv0.5.0Updated Oct 7, 2026

Every machine you have, for your coding agent: live GPU, VRAM, RAM and disk, and SSH.

VerifiedSTDIODesktop onlyCloud & InfrastructureDeveloper ToolsSecurity & Monitoring

Overview

AI-generated overview

Lets a coding agent see and use every machine you own — live GPU, VRAM, RAM and disk — and run commands over SSH.

What it does
fleet gives an assistant a single inventory of your machines, showing live GPU, VRAM, RAM and disk availability, and lets it run commands on any of them over SSH. It can match machines by capability (for example a free 24 GB card), report hourly cost and burn rate for paid rentals, and manage which machine may reach which. Agents with a shell get a skill describing the fleet commands; apps that cannot run commands get this MCP server, which runs the same commands for them.
When to use it
Worth adding when you work across several machines — local boxes, a NAS, or rented GPU instances — and want your agent to pick a free GPU, run a job on the right host, or check what is idle and costing money, instead of being limited to the machine it runs on.
Requirements
Runs locally as a stdio process, installed with uvx from the PyPI package agents-fleet; the README also shows a shell installer. The center machine needs an address it can route to (public IP, LAN, or an overlay such as Tailscale). Member machines need only sshd; SSH keys are used, with a password typed once by you if needed. No environment variables or headers are declared.
Before you install
The server runs commands on your other machines over SSH and can change SSH access between them, so an agent can act on remote hosts, not just read them. Credentials are not stored and the agent is not shown them, but you should still confirm which machine is meant before irreversible commands. Paid rentals are tracked with hourly prices and idle alerts; fleet does not launch new cloud VMs.

Installation

In SourceWeft

  1. Open fleet in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

fleet — let your coding agent use every machine you have

Claude Code, Codex and Gemini see one machine: the one they run on.
fleet shows them all of yours — every GPU, how much is free, and how to get there.

[tests] [License: MIT] [Linux | macOS | Windows] [MCP server]

[fleet ls: free GPU, VRAM, CPU, RAM and disk across seven machines — an A100 rental with one busy and one idle card, an idle H100 rental flagged as costing money, a busy RTX 3090 box with a nearly full disk, a laptop, a NAS and a machine that is switched off]

The problem

Ask your agent to train a model and it starts on your laptop — while a 4090 sits idle across the room and a rented A100 bills you by the hour. It cannot use what it cannot see.

  • Your agent is stuck on one machine. It has no idea your other boxes exist.
  • Finding a free GPU is manual. SSH into five hosts, run nvidia-smi, compare in your head.
  • Handing it a server means pasting credentials into the chat, and hoping.

Install once, then just talk to your agent

On the machine you work from:

bash
curl -LsSf https://raw.githubusercontent.com/lion-zhang/fleet/main/install.sh | sh

Windows: powershell -ExecutionPolicy ByPass -c "irm https://raw.githubusercontent.com/lion-zhang/fleet/main/install.ps1 | iex"

That is the whole setup. This machine becomes your fleet's center, and every supported agent installed on it learns fleet. From here on you say what you want in plain words — no commands to remember. When something is missing, the agent asks (illustrative):

You: add my new GPU server

Agent: Sure — how do you usually connect to it? An SSH command like ssh -p 40001 [email protected] is all I need.

You: ssh [email protected]

Agent: Added as gpu-box: 2× RTX 4090, both idle, 46 GB free. It's ready to use.

You: train train.py on whatever has a free 24 GB card

Agent: rtx4090 has 23.1 GB free and an idle GPU; a100-spot is free too but costs $1.89/hr. Starting on rtx4090, logging to train.log.

You sayWhat happens
"what's free right now?"every machine checked, the free ones listed
"find me a box with a 24 GB card"machines matched by what they have, not by name
"run the tests on the Linux box"run there, results brought back
"what's costing me money?"idle paid rentals flagged, with their hourly price
"let the laptop reach the NAS"access granted, applied at once
"add a machine without typing its password"a one-time line to paste there; it joins by itself

No hostnames, keys or passwords go into the conversation, and anything irreversible waits for your yes.

Every machine besides the center is a member. A member needs nothing installed — just sshd. For machines where you also want to run fleet, or that you would rather not type a password for, the agent gives you an invite line: pasted there, it installs fleet and joins by itself.

Already in your agent? Install from there

AgentInstall
Claude Code/plugin marketplace add lion-zhang/fleet then /plugin install fleet@fleet
Codexcodex plugin marketplace add lion-zhang/fleet then codex plugin add fleet@fleet
Gemini CLIgemini extensions install https://github.com/lion-zhang/fleet
GitHub Copilot CLIcopilot plugin marketplace add lion-zhang/fleet then copilot plugin install fleet@fleet
Cursor[Add to Cursor]
VS Code[Install in VS Code]
Claude Desktopopen fleet.mcpb from the latest release

Plus OpenCode, Amp, Windsurf, Cline, Zed, Qwen Code, Goose, Kiro, Hermes — 35+ agents in all: every agent →

On a machine in no fleet yet, these make it a center on first use, like the installer. For a member, paste its invite line first.

Built to be safe

  • Nothing to install on your machines. fleet probes with one script over one SSH connection — Linux, macOS or Windows. A NAS or a fresh rental works as it is.
  • Keys, not passwords. A password, if needed at all, is typed once by you. Nothing that could be stolen is stored, and the agent never sees a credential.
  • Access you control. The center decides which machine may reach which, and a revoke is pushed at once. fleet access gpu-box --allow laptop, --deny to take it back.
  • The agent asks, not guesses. It is told to ask which machine you mean, and to leave irreversible commands to you.

Prefer the command line?

Everything the agent does is a plain fleet command, for when you want to drive it yourself:

bash
fleet ls                            # every machine, with what is free right nowfleet ls --tag cuda --tag vram-24g  # by capability: NVIDIA, a card of 24 GB or morefleet top                           # live view, like htop for the whole fleetfleet show gpu-box                  # one machine in detail: GPU processes, services, disksfleet ssh gpu-box -- nvidia-smi     # run something therefleet add "ssh [email protected]"     # add a machine; fleet invite NAME for a join linefleet access nas --allow laptop     # let one machine reach another
[fleet top: a live view of GPU utilisation, free VRAM, CPU, RAM and disk across all machines]

Rentals from vast.ai, RunPod or Lambda show their price (fleet edit a100 --cost 1.89), the fleet's burn rate, and an alert when a paid machine sits idle. On Tailscale, ZeroTier or WireGuard? fleet just needs an address it can route to.

How fleet compares

fleet is about the machines you already have. It complements tools that launch new ones.

ssh + nvidia-sminvitop / gpustatSkyPilot / dstackfleet
All your machines in one view✗✗ one machine✓ the ones it manages✓
Built for coding agents (skill / MCP)✗✗partly✓
Nothing installed on target machines✓✗✗✓
Manages SSH access between machines✗✗for its own clusters✓
Launches new cloud VMs✗✗✓✗

FAQ

I use Claude Code and Codex (and more) on one machine. Does that work?

Yes — that is the normal case. fleet is installed once per machine: one command, one inventory, one set of keys. Each agent only gets a small skill or MCP entry pointing at it, so Claude Code, Codex, Gemini CLI and a desktop app all see the same machines, and can use them at the same time. Install a new agent later? Run fleet setup (or ask an agent that already has fleet to do it).

What does my agent actually get?

Agents with a shell (Claude Code, Codex, Gemini CLI, Copilot CLI, OpenCode, …) get a skill — text that tells them the fleet commands; it costs nothing until a task needs a machine. Apps that cannot run commands (Claude Desktop, Cursor, VS Code, …) get an MCP server that runs the same commands for them. The installer sets up the supported agents you have; docs/agents.md has the details for each.

Does it work behind NAT, or across sites?

The center needs an address it can route to: a public IP, a LAN, or an overlay such as Tailscale. A machine the center cannot dial can still join with the fleet invite line and report in; granting access to it waits until the center can reach it.

What if the center is off?

Normal — it can be a laptop that is closed half the day. Everything already granted keeps working; only changes wait for it. Move the role with fleet center NAME.

Windows?

Yes, both ways. A Windows machine works as a target with OpenSSH Server and nothing else, and fleet runs on Windows too, center included: interactive fleet ssh, fleet top and the background service all work there. See Windows.

Learn more

  • Getting started — the full walkthrough, every command
  • Every agent — install commands and config for 35+ agents
  • Design — one core per machine, a skill per agent, MCP for the rest; and access — how access is granted, signed and revoked

Status: v0.5. The test suite runs on Linux, macOS and Windows, and every command is run end to end on a real machine of each OS in CI — from a script, as an agent runs it, and at a real terminal, as you do. Multi-machine fleets (key, password, invite, handover) are tested on Linux machines built from scratch.

Issues and pull requests are welcome — uv run pytest -q runs the tests. If fleet saved you a GPU-hour, a ⭐ helps other people find it.

MIT

Source: README.md at commit 22afd65

Tools

0
Tool metadata has not been indexed yet.

Version history

1
  1. v0.5.0LatestOct 7, 2026