THE FIELD GUIDES

AI on a Windows PC

The same AI tools on a Windows laptop or desktop. Tap COPY and paste into Terminal.

On a Windows PC

The same tools on a Windows laptop or desktop. No installers to hunt for: each line installs one program, mostly with winget, which is built into Windows 10 and 11. Open Terminal (or PowerShell) from the Start menu and paste one line at a time. Close and reopen Terminal after installing, so new commands are found.

STEP 1
The basics: Git, Node, Bun and a newer PowerShell
winget install --id Git.Git -e
winget install --id OpenJS.NodeJS.LTS -e
powershell -c "irm bun.sh/install.ps1 | iex"
winget install --id Microsoft.PowerShell -e
Git saves and shares code, Node runs most AI coding tools, Bun is a faster JavaScript runtime that LifeOS needs (step 3). The old built-in PowerShell rejects && between commands; the new one (open it as "PowerShell 7") accepts it. On the old one, use ; instead.
STEP 2
Claude Code
npm install -g @anthropic-ai/claude-code
Then go to your project folder and start it. It opens the browser to sign in with your Claude account:
claude
STEP 3
LifeOS: give Claude Code a memory and your goals
LifeOS (free, by Daniel Miessler) runs on top of Claude Code. It remembers past sessions, keeps your goals in a TELOS file, and turns mistakes into rules so you don't repeat them. If you already use Claude Code, back up its settings first:
Copy-Item -Recurse $HOME\.claude $HOME\.claude-backup
Then start claude and paste this; Claude does the install and asks before changing anything:
Read https://ourlifeos.ai/install and install LifeOS for me.
Then fill in your TELOS (your mission, goals and projects) when it asks. That's what makes it yours. Want the files themselves? It's the LifeOS ZIP in the list further down.
STEP 4
Codex, Gemini CLI and OpenCode
npm install -g @openai/codex
npm install -g @google/gemini-cli
npm install -g opencode-ai
Start each with codex, gemini or opencode, and sign in the way it asks: ChatGPT for Codex, Google for Gemini, /connect in OpenCode.
STEP 5
Ollama: free models on your own PC
winget install --id Ollama.Ollama -e
Download and chat with a model:
ollama run gemma3:4b
Or run a coding tool on it (the tool must be installed first):
ollama launch claude
STEP 6
Use your PC's Ollama in ChathoNU
Let chathonu.com talk to Ollama, then quit Ollama from the tray icon and open it again:
setx OLLAMA_ORIGINS "https://chathonu.com"
Then open chathonu.com/local, choose MY OLLAMA and connect to:
http://localhost:11434
STEP 7
GitHub and Vercel (to publish what you build)
winget install --id GitHub.cli -e
gh auth login
npm install -g vercel
vercel login
STEP 8
Optional: phone access, Linux, Docker and Python
Tailscale puts your phone and PC on one private network, so your phone can use the PC's Ollama from anywhere (see below):
winget install --id Tailscale.Tailscale -e
Want a real Linux terminal on Windows too? This installs Ubuntu inside Windows (restart when asked):
wsl --install
Docker runs the self-hosted projects below (MCP servers, n8n, Frigate). Python runs A.D.A.:
winget install --id Docker.DockerDesktop -e
winget install --id Python.Python.3.12 -e

The projects I use (free downloads)

These are the ZIP files behind my setup. Each is someone else's open project, so download it from the original page to get the latest version: open the link, tap the green Code button, then Download ZIP. Or clone it with Git, for example:

git clone https://github.com/danielmiessler/LifeOS
LifeOS
github.com/danielmiessler/LifeOS
The core of my setup: a life operating system for Claude Code with memory, goals (TELOS) and skills. Easiest install is the one-line prompt in step 3.
Termux
github.com/termux/termux-app
The Linux terminal on my phone that runs everything in the Android guides. Install the app from F-Droid or this page's Releases, not the Play Store.
Remote Claude Code (NetworkChuck)
github.com/theNetworkChuck/remote-claude-code
Run Claude Code on a cheap server and reach it from your phone over SSH, so heavy work doesn't run on the phone.
Docker MCP tutorial (NetworkChuck)
github.com/theNetworkChuck/docker-mcp-tutorial
Build MCP servers in Docker, which are the tools that let Claude and other AIs reach your apps and files. Needs Docker Desktop.
n8n AI agent "Terry" (NetworkChuck)
github.com/theNetworkChuck/n8n-terry-guide
An n8n automation agent that watches a home server and fixes problems with your approval. Needs Docker.
Frigate NVR guide (NetworkChuck)
github.com/theNetworkChuck/frigate-nvr-guide
A fully local AI security camera system, nothing sent to the cloud. Needs Docker; runs on a PC or Raspberry Pi.
A.D.A. voice assistant
github.com/nazirlouis/ada
A real-time talking assistant built on Gemini Live, with webcam and screen sharing. Needs Python and a Gemini API key.
AI hacking study prompts (NetworkChuck)
github.com/theNetworkChuck/ai_hacking_study_prompts
Ready-made prompts that turn an AI into a tutor for studying ethical hacking (HTB CPTS). Just copy and paste.
Goof (Snyk)
github.com/snyk/goof
A deliberately insecure app for practicing finding and fixing security bugs. Run it only on your own PC, never on a server anyone can reach.

Use your PC's Ollama from your phone

STEP 1
Give it an https address
Browsers block plain http from a secure site. The easy way is Tailscale (step 8). On the PC run:
tailscale serve --bg 11434
and use the https address it prints in MY OLLAMA on chathonu.com/local.

Stuck on a step? Send a message and say which one.