← OSWork area

OSWork Training · 3 courses in order

From the chat to your agent environment, command by command.

The terminal is just another window onto the same Windows folders. With six commands you move through them, with Claude Code or Codex the agent works inside the folder, with Git you save each version, and on a VPS the same project runs 24 hours a day. In 3 courses you do all of this, from your first cmd to your AI operating system.

Start with OSWork Quick See what you will learn

To get started: a computer with Windows 10 or 11 and a Claude or ChatGPT subscription.

OSWork — a 3-course program: from chat to your own agent environment, command by command.
3courses, from the first command to the AI OS
71lessons and modules in courses 2 and 3
17command-by-command steps in the area
3languages: PT · EN · ES

The problem

The AI answers well. The work stays in the chat.

What separates people who use AI from people who work with agents is the environment: folder, terminal, versions and a server. Each course builds one part of it.

  • You copy the answer from the chat and paste it into your files, one by one.
  • The terminal looks dangerous, and you don't know where things are.
  • An agent mistake wipes out what you had, and there is no version to go back to.
  • Your project only exists on your computer: it is not on GitHub and not online.
  • Everything stops when the computer shuts down, because you have never used a VPS.

After the training

By the end of the 3 courses, you can:

01

Move around the terminal without fear

cmd and folders on Windows, six navigation commands and File Explorer showing the same place.

02

Put the agent to work in the folder

Claude Code or Codex installed from the official source, AGENTS.md with the project rules and reusable Skills.

03

Save and publish each version

Git, gh and an SSH key on GitHub, the everyday routine and your page on GitHub Pages.

04

Run the same project on a VPS

Log in over SSH, install only what is needed, protect the network and credentials, and supervise with systemd.

05

Build your AI operating system

The 6 layers, identity in CLAUDE.md, rules and hooks, skills, tools and agents, with an audit against the goal.

3 courses in order · 71 lessons and modules

Three courses, from the first cmd to the system that works for you.

Quick is the short path, with the same steps as the area in four parts. v6.2 explains every step from scratch. The AI OS engineering course takes the ready environment to a layered system.

  1. 1

    Short path

    OSWork Quick

    In four parts: terminal, folders and agent on Windows, Git with gh and an SSH key, the VPS and an overview of Vercel and Supabase, with copy-and-paste commands.

    Open course 1
  2. 2

    Step by step

    OSWork v6.2

    Eight modules of short lessons: models, Chat, Work and Desktop, terminal and Codex, folders and secrets, AGENTS and Skills, Git, a Telegram bot and a 24/7 VPS.

    49 lessonsOpen course 2
  3. 3

    Advanced

    AI Operating System Engineering

    The 6 layers of an AI OS, the step-by-step with /os-coach through to the audit, and real OSes for sales, support, content and consulting.

    22 modulesOpen course 3

Who it's for

Made for you if

  • You use AI only in the chat and want it to work inside your folders.
  • You have never opened the terminal, or you opened it and didn't know what to type.
  • You want your project on GitHub, published and with every version saved.
  • You want a server running 24 hours a day and don't know where to start.

Not for you yet if

  • You already work in the terminal every day: go to the From chat to agent area or to the JARVIS area.
  • You want to skip the practice and just watch.

Included in the training

Each course has an open repository and a video.

The courses are on INEMA's GitHub, with copy-and-paste commands and video lessons featuring Nei's avatar and voice.

Projects and kits

OS Coachoswork-v62 · GitHuboswork-quick · GitHub

Videos

The step-by-step, command by commandOSWork Quick: the seven lessonsOSWork v6.2: video lessons30-second short

Why this training

What changes compared with learning on your own.

Command and what you see

Each step shows the command to copy and what should appear on screen. You know right away whether it worked.

Windows and VPS the same way

If you learn to navigate on Windows, you navigate on the VPS the same way: only a few names change.

Two paces to start

Quick is for those who want to go straight to the point; v6.2, with short lessons and a glossary, is for those who want every step explained.

Three languages, same content

All three courses exist in Portuguese, English and Spanish.

Questions

Before you start.

Do I need to know how to code?

No. You copy and paste the commands, and the agent writes the code inside the project folder.

Quick or v6.2?

Quick covers the same steps as the area, in four parts. v6.2 explains every step from scratch in 49 short lessons, with a glossary. Do one of the two, then course 3.

Which tools do I use?

Windows cmd, Git, GitHub CLI, Node, Claude Code or Codex, GitHub Pages and an Ubuntu VPS.

How long does it take?

Quick has four parts. v6.2 has 49 lessons in eight modules. AI OS engineering has 22 modules in five tracks.

Do I have to pay for the VPS?

The VPS is rented separately from a provider. Parts 1 and 2 work only on your own computer and on GitHub.

Open the terminal today and create your projects folder.

Start with OSWork Quick. In four parts, you go from an empty terminal to a published website and the same project running on a VPS.

Start with OSWork Quick See the OSWork area