Nixos Managing logo

Nixos Managing

Community
michalzubkowicz
nixos-managing

Use when managing NixOS systems — rebuilding, configuring, deploying, installing, or building images. Covers flakes, modules, secret management, VM management, disk imaging, remote deployment, and common anti-patterns to avoid.

Overview

Publishermichalzubkowicz
Repositorynixos-management-skill
Skill namenixos-managing
Stars
67
Forks
1
Bundled files
8
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 8 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by michalzubkowicz on GitHub. Read the source before you install it.

Installation

Install the Nixos Managing AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/michalzubkowicz/nixos-management-skill/tree/main/nixos-managing
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/michalzubkowicz/nixos-management-skill.git /tmp/nixos-management-skill
mkdir -p .claude/skills
cp -r /tmp/nixos-management-skill/nixos-managing .claude/skills/nixos-managing
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Nixos Managing in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Nixos Managing on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Nixos Managing is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

NixOS Management

Quick Decision: What do you need?

TaskGo to
Rebuild / activate / rollback systemvm-management.md
Install NixOS on new machineinstallation.md
Build ISO / disk image / VM imageimage-building.md
Configure modules, flakes, packages, servicesconfiguration.md
Ephemeral root / wipe on boot / impermanenceimpermanence.md
LUKS encryption / remote unlock (SSH, Tailscale)luks.md
Health checks / Telegram (or webhook) alerts for filesystem, disk, SMART, servicesmonitoring.md
Something isn't working / weird behavioranti-patterns.md

Execution Context — Ask First

Before suggesting any commands, establish where they will run.

NixOS management involves multiple machines. Commands that work on the NixOS host will fail on macOS or non-NixOS Linux. Always determine the setup before suggesting commands.

If unsure — ask: "Where are you editing configs, and where does nixos-rebuild run?"

All common setups

Setup A — NixOS local (everything on one machine)

NixOS machine: edit → nixos-rebuild switch (local)
  • All commands run locally
  • nixos-option, man configuration.nix, nixos-rebuild — all local

Setup B — NixOS workstation → NixOS remote server

NixOS workstation (edit) → nixos-rebuild --target-host → NixOS server
  • nixos-rebuild runs on workstation, deploys to server
  • Option verification: workstation (has nixpkgs) or server via SSH
  • nixos-option on server reflects what's actually active there

Setup C — macOS → NixOS remote (Linux rebuilds locally)

macOS (edit .nix files) → push/rsync → NixOS Linux host → nixos-rebuild switch (local)
  • nixos-rebuild runs on the Linux host (SSH in, or via remote trigger)
  • Option verification: SSH into Linux host — macOS has no nixos-option
  • search.nixos.org/options always works from macOS

Setup D — macOS → NixOS remote (macOS drives rebuild)

macOS (edit) → nixos-rebuild --target-host root@linux-host --flake .#host
  • nixos-rebuild runs on macOS but activates on Linux
  • Requires Nix installed on macOS (nix CLI, not NixOS)
  • Option verification: nix eval works on macOS if Nix is installed; otherwise use search.nixos.org

Setup E — Linux (non-NixOS) → NixOS remote

Ubuntu/Debian/Arch (edit + run nixos-rebuild) → NixOS remote server
  • Same as Setup D — Nix must be installed on the source machine
  • nixos-option not available unless NixOS; use nix eval or search.nixos.org

Setup F — CI/CD → NixOS remote

CI agent (GitHub Actions / GitLab CI) → deploy to NixOS server
  • CI agent needs Nix installed (use cachix/install-nix-action or similar)
  • No interactive verification — all options must be pre-validated
  • Use nix flake check in CI to catch errors before deploy

Verification availability by machine type

Machinenixos-optionnix eval nixpkgs#...man configuration.nixsearch.nixos.org
NixOS host✅✅✅✅
macOS with Nix❌✅❌✅
Linux (non-NixOS) with Nix❌✅❌✅
macOS without Nix❌❌❌✅
CI agent with Nix❌✅❌✅

MCP Server: mcp-nixos (optional accelerator)

If the mcp-nixos MCP server (utensils/mcp-nixos) is configured in the environment, prefer it for lookups over SSH/web round-trips — but ask the user first before using it for a given task: "I can use the mcp-nixos MCP server to look this up directly — want me to use it?"

What it's good at:

  • Exact NixOS package names, versions, metadata (130K+ packages)
  • NixOS options (23K+) — verify option paths before suggesting config
  • Home Manager options (5K+)
  • nix-darwin macOS settings (1K+)
  • Nixvim Neovim configuration (5K+)
  • FlakeHub registry discovery (600+ flakes)
  • Noogle Nix function lookup (2K+)
  • NixOS Wiki / nix.dev documentation
  • Binary cache status checks
  • Local flake input exploration
  • Package version history with nixpkgs commit hashes (reproducible pins)

Use it instead of: guessing option names, scraping search.nixos.org, or SSHing into a host just to run nixos-option. Works on macOS without Nix installed.

Tools exposed: nix (unified query), nix_versions (version history).

When NOT to ask: the user has already said they want (or don't want) to use it this session — honor that preference.

Before installing — security review required. Any MCP server runs with your shell privileges. Before adding mcp-nixos (or any MCP) to the configuration:

  1. Download the source (git clone https://github.com/utensils/mcp-nixos or inspect the PyPI tarball via uvx --from mcp-nixos --print / pip download mcp-nixos --no-deps).
  2. Review it for anything harmful: unexpected network endpoints, filesystem writes outside its own cache, shell/exec calls, credential reads (~/.ssh, ~/.aws, env vars), obfuscated code, suspicious post-install hooks in pyproject.toml / setup.py.
  3. Pin to a known-good version/commit rather than a floating tag.
  4. Only after review — add it to the MCP config.

Tell the user what you reviewed and what you found before they approve the install.

Option Verification — Always Verify Before Suggesting

Never suggest a NixOS option from memory alone. Option names change between NixOS versions, and incorrect options fail silently or produce confusing errors.

Fastest path (if available): the mcp-nixos MCP server — see section above. Ask the user before using it.

Run on the NixOS host (not macOS, not CI unless it runs NixOS):

bash
# Search available options interactively
nixos-option services.openssh.settings

# Evaluate option existence in current nixpkgs
nix eval nixpkgs#nixosOptionsDoc --apply 'x: builtins.attrNames x' 2>/dev/null | grep -o '"[^"]*PasswordAuth[^"]*"'

# Quickest: man page on the NixOS machine
man configuration.nix | grep -A3 "PasswordAuthentication"

Or use the web (version-specific, always available):

  • https://search.nixos.org/options — search by option name, filterable by channel
  • Check the channel matching your nixpkgs.url (e.g. nixos-26.05)

Workflow when unsure about an option:

  1. Check search.nixos.org/options for the option name and correct path
  2. Note the NixOS version it applies to
  3. Only then include it in configuration

Core Mental Model

NixOS is declarative and atomic. Every change produces a new generation. You can always roll back.

Key workflow:

  1. Edit .nix files
  2. git add (in flakes — untracked files are invisible to Nix)
  3. nixos-rebuild test (activate without committing to bootloader)
  4. nixos-rebuild switch (set as default boot)

Flakes vs Channels

Use flakes for reproducibility. Channels are impure (machine-dependent lookup paths).

Enable flakes once:

nix
nix.settings.experimental-features = [ "nix-command" "flakes" ];

Common nixos-rebuild Commands

CommandWhen to use
nixos-rebuild switchApply + set as default boot
nixos-rebuild testApply now, skip bootloader — safe first step
nixos-rebuild bootSet as next boot without activating now
nixos-rebuild dry-activatePreview changes without applying
nixos-rebuild buildBuild only, creates ./result
nixos-rebuild build-vmBuild QEMU VM for local testing

Always test before switch — especially on remote servers:

bash
nixos-rebuild test --flake .#hostname --target-host root@server
nixos-rebuild switch --flake .#hostname --target-host root@server

Remote Deployment Tools

ToolBest for
nixos-rebuild --target-host1-3 machines, simplest
deploy-rsSmall fleet, auto-rollback on failure
colmenaLarge fleet, parallel, cross-host config
nixos-anywhereInitial install on non-NixOS machine

Secret Management

ToolWhen to use
agenixSimple setup, SSH key workflow, small secret count
sops-nixCloud KMS needed, templating, multiple formats

Secrets live at /run/agenix/ or /run/secrets/ — never in Nix store.

Most Frequently Changed Elements

  • environment.systemPackages — installed packages
  • users.users.<name> — user accounts and SSH keys
  • networking.* — hostname, IPs, firewall ports
  • services.* — enable/configure systemd services
  • boot.loader.* — bootloader and kernel settings
  • nix.settings.* — substituters, trusted users, features

See configuration.md for patterns.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Nixos Managing AI skill do?

Use when managing NixOS systems — rebuilding, configuring, deploying, installing, or building images. Covers flakes, modules, secret management, VM management, disk imaging, remote deployment, and common anti-patterns to avoid.

Why use Nixos Managing on TypingMind?

Because you install it once and use it with any model. Nixos Managing is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Nixos Managing in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/michalzubkowicz/nixos-management-skill/tree/main/nixos-managing. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Nixos Managing?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Nixos Managing?

As many as you like. As long as a model supports skills, you can use Nixos Managing with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Nixos Managing AI skill free?

Yes. It is published on GitHub by michalzubkowicz under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇