agents-fleet 0.5.0__tar.gz

agents_fleet-0.5.0/.claude-plugin/marketplace.json

@@ -0,0 +1,35 @@
+{
+  "name": "fleet",
+  "owner": {
+    "name": "Lin Zhang",
+    "url": "https://github.com/lion-zhang"
+  },
+  "metadata": {
+    "description": "fleet, as a Claude Code plugin."
+  },
+  "plugins": [
+    {
+      "name": "fleet",
+      "source": "./",
+      "description": "Every machine you have, for your coding agent: live GPU, VRAM, RAM and disk, and SSH.",
+      "version": "0.5.0",
+      "author": {
+        "name": "Lin Zhang",
+        "url": "https://github.com/lion-zhang"
+      },
+      "homepage": "https://github.com/lion-zhang/fleet",
+      "license": "MIT",
+      "keywords": [
+        "gpu",
+        "inventory",
+        "ssh",
+        "remote",
+        "compute",
+        "mcp",
+        "homelab",
+        "mlops"
+      ],
+      "category": "devops"
+    }
+  ]
+}

agents_fleet-0.5.0/.claude-plugin/plugin.json

@@ -0,0 +1,22 @@
+{
+  "name": "fleet",
+  "version": "0.5.0",
+  "description": "Every machine you have, for your coding agent: live GPU, VRAM, RAM and disk, and SSH.",
+  "author": {
+    "name": "Lin Zhang",
+    "url": "https://github.com/lion-zhang"
+  },
+  "homepage": "https://github.com/lion-zhang/fleet",
+  "repository": "https://github.com/lion-zhang/fleet",
+  "license": "MIT",
+  "keywords": [
+    "gpu",
+    "inventory",
+    "ssh",
+    "remote",
+    "compute",
+    "mcp",
+    "homelab",
+    "mlops"
+  ]
+}

agents_fleet-0.5.0/.gitattributes

@@ -0,0 +1,14 @@
+# The probe payloads are sent verbatim over ssh to a shell on the far side, so their
+# line endings are a wire format, not a local preference. Git for Windows defaults to
+# core.autocrlf=true, which rewrote payload.sh to CRLF on checkout -- and every probe
+# from the Windows center then failed with `sh: 8: \r: not found` followed by
+# `Syntax error: "|" unexpected`, because dash reads the CR as part of the command.
+#
+# Declared here rather than fixed on the machine: it is a property of the file, and the
+# next clone on any Windows box would otherwise reintroduce it.
+*.sh text eol=lf
+*.ps1 text eol=lf
+
+# Same reasoning, one layer up: these are read and piped to a remote shell too.
+src/fleet/probe/payload.sh text eol=lf
+src/fleet/probe/payload.ps1 text eol=lf

agents_fleet-0.5.0/.github/workflows/e2e.yml

@@ -0,0 +1,85 @@
+name: e2e
+
+# fleet installed and used for real on each OS: the installer makes the runner a center,
+# then tests/e2e/host.py runs every command against it -- over ssh to the runner itself
+# as well, so the probe, enrolment and authorized_keys edits go through that OS's sshd
+# and shell. Slower than the unit tests, so only for changes that can break it.
+on:
+  workflow_dispatch:
+  pull_request:
+    paths: ["src/**", "install.sh", "install.ps1", "tests/e2e/**", ".github/workflows/e2e.yml"]
+  push:
+    branches: [main]
+    paths: ["src/**", "install.sh", "install.ps1", "tests/e2e/**", ".github/workflows/e2e.yml"]
+
+jobs:
+  e2e:
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest, macos-latest, windows-latest]
+    runs-on: ${{ matrix.os }}
+    timeout-minutes: 30
+    steps:
+      - uses: actions/checkout@v4
+      - uses: astral-sh/setup-uv@v6
+        with:
+          python-version: "3.12"
+
+      - name: sshd (Linux)
+        if: runner.os == 'Linux'
+        run: |
+          sudo apt-get install -y -qq openssh-server >/dev/null
+          sudo systemctl start ssh
+      - name: sshd (macOS)
+        if: runner.os == 'macOS'
+        run: |
+          sudo launchctl load -w /System/Library/LaunchDaemons/ssh.plist 2>/dev/null \
+            || sudo systemsetup -f -setremotelogin on
+          for i in $(seq 1 20); do nc -z localhost 22 && break; sleep 1; done
+      # sshd refuses a key when the home or ~/.ssh is writable by others (StrictModes);
+      # prove a key login works before blaming fleet for one that does not.
+      - name: a key login to localhost works
+        if: runner.os != 'Windows'
+        run: |
+          chmod go-w "$HOME"; mkdir -p ~/.ssh; chmod 700 ~/.ssh
+          ssh-keygen -q -t ed25519 -N "" -f ~/.ssh/e2e_check
+          cat ~/.ssh/e2e_check.pub >> ~/.ssh/authorized_keys; chmod 600 ~/.ssh/authorized_keys
+          ls -ld "$HOME" ~/.ssh ~/.ssh/authorized_keys
+          ssh -i ~/.ssh/e2e_check -o BatchMode=yes -o StrictHostKeyChecking=accept-new localhost true \
+            || { ssh -vv -i ~/.ssh/e2e_check -o BatchMode=yes localhost true 2>&1 | tail -30; exit 1; }
+      - name: sshd (Windows)
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          if (-not (Get-Service sshd -ErrorAction SilentlyContinue)) {
+            Add-WindowsCapability -Online -Name OpenSSH.Server~~~~0.0.1.0 | Out-Null
+          }
+          Start-Service sshd
+          Get-Service sshd
+
+      - name: install.sh (center)
+        if: runner.os != 'Windows'
+        run: |
+          FLEET_SOURCE="$PWD" sh install.sh
+          echo "$HOME/.local/bin" >> "$GITHUB_PATH"
+      - name: install.ps1 (center)
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          $env:FLEET_SOURCE = "$PWD"
+          ./install.ps1
+          uv tool dir --bin | Out-File -Append -Encoding utf8 $env:GITHUB_PATH
+
+      - name: every command
+        if: runner.os != 'Windows'
+        run: |
+          FLEET_BIN="$HOME/.local/bin/fleet" uv run --no-project --with mcp --with . \
+            python tests/e2e/host.py --ssh-user "$USER" --report "$GITHUB_STEP_SUMMARY"
+      - name: every command
+        if: runner.os == 'Windows'
+        shell: pwsh
+        run: |
+          $env:FLEET_BIN = Join-Path (uv tool dir --bin) 'fleet.exe'
+          # pywinpty: a ConPTY pseudo-console, to type into fleet as a person would
+          uv run --no-project --with mcp --with pywinpty --with . python tests/e2e/host.py --ssh-user $env:USERNAME --report $env:GITHUB_STEP_SUMMARY

agents_fleet-0.5.0/.github/workflows/release.yml

@@ -0,0 +1,67 @@
+name: release
+
+# Publishing a GitHub release (tag vX.Y.Z) puts that version on PyPI as agents-fleet and
+# attaches the wheel, the sdist and the Claude Desktop bundle (fleet.mcpb) to it.
+#
+# PyPI is reached by Trusted Publishing: no token is stored anywhere. pypi.org must list
+# this repository, this file and the `pypi` environment as a publisher of agents-fleet.
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: astral-sh/setup-uv@v6
+      - name: The tag is the version being released
+        run: |
+          v="$(uv version --short)"
+          test "v$v" = "$GITHUB_REF_NAME" || { echo "tag $GITHUB_REF_NAME, pyproject.toml $v"; exit 1; }
+      - name: The generated plugin, extension and registry files are current
+        run: |
+          uv run python scripts/build_dist.py
+          git diff --exit-code
+      - run: uv build
+      - uses: actions/setup-node@v4
+        with:
+          node-version: 22
+      - name: Claude Desktop bundle
+        run: |
+          npx -y @anthropic-ai/mcpb validate mcpb/manifest.json
+          npx -y @anthropic-ai/mcpb pack mcpb dist/fleet.mcpb
+      - uses: actions/upload-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+  pypi:
+    needs: build
+    runs-on: ubuntu-latest
+    environment: pypi
+    permissions:
+      id-token: write
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+      # The bundle is for the release page, not for PyPI.
+      - run: rm dist/fleet.mcpb
+      - uses: pypa/gh-action-pypi-publish@release/v1
+
+  assets:
+    needs: build
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+      - run: gh release upload "$GITHUB_REF_NAME" dist/* --clobber --repo "$GITHUB_REPOSITORY"
+        env:
+          GH_TOKEN: ${{ github.token }}

agents_fleet-0.5.0/.github/workflows/test.yml

@@ -0,0 +1,27 @@
+name: tests
+
+on:
+  # A branch with a pull request is tested by the pull_request run; testing its pushes
+  # too ran every commit twice and sent every failure twice.
+  push:
+    branches: [main]
+  pull_request:
+  workflow_dispatch:
+
+jobs:
+  test:
+    strategy:
+      fail-fast: false
+      matrix:
+        # Every OS fleet runs on: the user's Windows machine found what Linux and macOS
+        # runs could not (top's termios, ssh's exec), so Windows runs the suite too.
+        os: [ubuntu-latest, macos-latest, windows-latest]
+        python: ["3.12", "3.13"]
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v4
+      - uses: astral-sh/setup-uv@v6
+        with:
+          python-version: ${{ matrix.python }}
+      - run: uv sync --all-extras
+      - run: uv run pytest -q

agents_fleet-0.5.0/.gitignore

@@ -0,0 +1,20 @@
+__pycache__/
+*.pyc
+.venv/
+dist/
+build/
+*.egg-info/
+.pytest_cache/
+
+# Runtime state -- lives in the OS config/state dirs, never in the repo.
+# Listed defensively in case it is ever copied here for debugging.
+inventory.yaml
+config.yaml
+secrets.age
+*.db
+*.db-wal
+*.db-shm
+
+# Local data / scratch -- never pushed.
+data/
+.remember/

agents_fleet-0.5.0/AGENTS.md

@@ -0,0 +1,4 @@
+# fleet — notes for agents
+
+Read [CLAUDE.md](CLAUDE.md) first: it holds the design decisions for this repo (the three
+layers — core, skill, MCP — and the write-queue rule), and how to work here.

agents_fleet-0.5.0/CLAUDE.md

@@ -0,0 +1,30 @@
+# fleet — notes for agents working on this repo
+
+## Design decisions — do not revisit
+
+fleet has three layers (full text: [docs/design/layers.md](docs/design/layers.md)):
+
+- **Core** — the `fleet` CLI plus its state and configuration, installed once per
+  machine. People can always run it by hand.
+- **Skill** — a prompt telling an agent how to use the CLI properly. Text only.
+- **MCP** — runs the same CLI for agents that cannot run commands. A thin wrapper,
+  never a second implementation.
+
+Consequences:
+- Add a capability to the CLI first; the skill describes it, MCP calls the CLI for it.
+- Reads run in parallel and never wait; state files are only replaced atomically.
+- Writes go through one queue and apply to the state as it is now — never save a copy
+  loaded earlier. Nobody blocks on anybody.
+- The installer never overwrites an existing core or touches fleet's state.
+
+## Working here
+
+- Tests: `uv run pytest -q`; CI runs them on Linux, macOS and Windows.
+- End to end: `tests/e2e/host.py` runs every command against a real install, on each OS
+  (`.github/workflows/e2e.yml`, also from the Actions tab). It drives `fleet top` and
+  `fleet ssh` at a real terminal (`tests/e2e/term.py`: a pty, or ConPTY on Windows) —
+  agents run fleet with no terminal, people with one, and both must be tested.
+- Generated files, never hand-edited: `skills/fleet/SKILL.md` and the plugin/extension
+  manifests (`uv run python scripts/build_dist.py`), the README screenshots
+  (`uv run python scripts/readme_screenshots.py`).
+- The two install modes are **center** and **member**; use those words.

agents_fleet-0.5.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Lin Zhang
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

agents_fleet-0.5.0/PKG-INFO

@@ -0,0 +1,232 @@
+Metadata-Version: 2.5
+Name: agents-fleet
+Version: 0.5.0
+Summary: Give Claude Code, Codex and Gemini every machine you have: a live GPU, VRAM, RAM and disk inventory across your servers, SSH access between them, and an MCP server.
+Project-URL: Homepage, https://github.com/lion-zhang/fleet
+Project-URL: Documentation, https://github.com/lion-zhang/fleet/blob/main/docs/getting-started.md
+Project-URL: Issues, https://github.com/lion-zhang/fleet/issues
+Author: Lin Zhang
+License-Expression: MIT
+License-File: LICENSE
+Keywords: ai-agents,claude-code,codex,coding-agent,devops,gemini-cli,gpu,gpu-monitoring,homelab,inventory,mcp,mcp-server,mlops,nvidia,remote-execution,runpod,ssh,ssh-keys,vast-ai,vram
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: Intended Audience :: System Administrators
+Classifier: Operating System :: MacOS
+Classifier: Operating System :: Microsoft :: Windows
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: System :: Clustering
+Classifier: Topic :: System :: Monitoring
+Classifier: Topic :: System :: Systems Administration
+Requires-Python: >=3.12
+Requires-Dist: filelock>=3.13
+Requires-Dist: mcp>=2
+Requires-Dist: platformdirs>=4.2
+Requires-Dist: psutil>=5.9
+Requires-Dist: pyyaml>=6.0
+Requires-Dist: rich>=13.7
+Requires-Dist: typer>=0.12
+Provides-Extra: mcp
+Provides-Extra: migrate
+Requires-Dist: pyrage>=1.1; extra == 'migrate'
+Description-Content-Type: text/markdown
+
+<!-- mcp-name: io.github.lion-zhang/fleet -->
+<div align="center">
+
+# fleet — let your coding agent use every machine you have
+
+**Claude Code, Codex and Gemini see one machine: the one they run on.<br>
+fleet shows them all of yours — every GPU, how much is free, and how to get there.**
+
+[![tests](https://github.com/lion-zhang/fleet/actions/workflows/test.yml/badge.svg)](https://github.com/lion-zhang/fleet/actions/workflows/test.yml)
+[![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)
+![Linux | macOS | Windows](https://img.shields.io/badge/platform-linux%20%7C%20macos%20%7C%20windows-lightgrey.svg)
+![MCP server](https://img.shields.io/badge/MCP-server-8A2BE2.svg)
+
+<img src="docs/assets/fleet-ls.svg" alt="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" width="100%">
+
+</div>
+
+## 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
+```
+
+<sub>Windows: `powershell -ExecutionPolicy ByPass -c "irm https://raw.githubusercontent.com/lion-zhang/fleet/main/install.ps1 | iex"`</sub>
+
+That is the whole setup. This machine becomes your fleet's **center**, and every
+[supported agent](docs/agents.md) 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 root@1.2.3.4` is all I need.
+>
+> **You:** `ssh ubuntu@10.0.0.7`
+>
+> **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 say | What 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
+
+| Agent | Install |
+|---|---|
+| **Claude Code** | `/plugin marketplace add lion-zhang/fleet` then `/plugin install fleet@fleet` |
+| **Codex** | `codex plugin marketplace add lion-zhang/fleet` then `codex plugin add fleet@fleet` |
+| **Gemini CLI** | `gemini extensions install https://github.com/lion-zhang/fleet` |
+| **GitHub Copilot CLI** | `copilot plugin marketplace add lion-zhang/fleet` then `copilot plugin install fleet@fleet` |
+| **Cursor** | [![Add to Cursor](https://cursor.com/deeplink/mcp-install-dark.svg)](https://cursor.com/en/install-mcp?name=fleet&config=eyJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJhZ2VudC1mbGVldCIsIm1jcCJdfQ%3D%3D) |
+| **VS Code** | [![Install in VS Code](https://img.shields.io/badge/VS_Code-Install_fleet-0098FF?logo=visualstudiocode&logoColor=white)](https://insiders.vscode.dev/redirect/mcp/install?name=fleet&config=%7B%22command%22%3A%22uvx%22%2C%22args%22%3A%5B%22agents-fleet%22%2C%22mcp%22%5D%7D) |
+| **Claude Desktop** | open `fleet.mcpb` from the [latest release](https://github.com/lion-zhang/fleet/releases/latest) |
+
+Plus OpenCode, Amp, Windsurf, Cline, Zed, Qwen Code, Goose, Kiro, Hermes — 35+ agents in all:
+**[every agent →](docs/agents.md)**
+
+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 now
+fleet ls --tag cuda --tag vram-24g  # by capability: NVIDIA, a card of 24 GB or more
+fleet top                           # live view, like htop for the whole fleet
+fleet show gpu-box                  # one machine in detail: GPU processes, services, disks
+fleet ssh gpu-box -- nvidia-smi     # run something there
+fleet add "ssh ubuntu@10.0.0.7"     # add a machine; fleet invite NAME for a join line
+fleet access nas --allow laptop     # let one machine reach another
+```
+
+<img src="docs/assets/fleet-top.svg" alt="fleet top: a live view of GPU utilisation, free VRAM, CPU, RAM and disk across all machines" width="100%">
+
+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-smi` | nvitop / gpustat | SkyPilot / dstack | **fleet** |
+|---|:---:|:---:|:---:|:---:|
+| 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
+
+<details>
+<summary><b>I use Claude Code <i>and</i> Codex (and more) on one machine. Does that work?</b></summary>
+
+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).
+</details>
+
+<details>
+<summary><b>What does my agent actually get?</b></summary>
+
+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](docs/agents.md) has the details for each.
+</details>
+
+<details>
+<summary><b>Does it work behind NAT, or across sites?</b></summary>
+
+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.
+</details>
+
+<details>
+<summary><b>What if the center is off?</b></summary>
+
+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`.
+</details>
+
+<details>
+<summary><b>Windows?</b></summary>
+
+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](docs/getting-started.md#windows).
+</details>
+
+## Learn more
+
+- [Getting started](docs/getting-started.md) — the full walkthrough, every command
+- [Every agent](docs/agents.md) — install commands and config for 35+ agents
+- [Design](docs/design/layers.md) — one core per machine, a skill per agent, MCP for the rest; and [access](docs/design/access.md) — 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](LICENSE)