OSWork
AI as a work system

From chat to your agent workspace. Learn to organize files, instructions, memory and tools into a work system you can verify.

Editions: technical v2, the full step-by-step v6.2, v6 and v5 for beginners, and Quick. Courses and videos in Portuguese, Spanish and English.

The banner presents the technical v2 path. Explore the v5 and Quick editions below as well.

Choose your OSWork edition

Five paths to organize your work with AI. Choose by your experience and the time you have.

OSWork v2 — technical path

Eight modules, four tracks and 48 topics. From chat to the terminal and Codex, files, Skills, Git, a Telegram bot and supervised VPS operation. Includes a practice kit.

OSWork v6.2 — the full path, step by step

The complete v2 path rebuilt for beginners: eight modules and 48 lessons of about 15 minutes. From chat to the terminal and Codex, folders, Skills, Git, a Telegram bot and a VPS, with visual step-by-steps and a glossary of technical terms.

OSWork v6 — beginners, visual

For people just starting out: seven lessons of about 15 minutes with simulated chat screens, real before-and-after examples and visual step-by-steps. Build folders per topic, a sheet with your rules, a request template and a checking routine, with nothing to set up.

OSWork v5 — for beginners

For professionals aged 40+ with little AI experience. Seven lessons of 20–25 minutes to organize folders, instructions, reusable requests and a review routine, without programming.

OSWork Quick — straight to the point

A short, visual edition for professionals aged 40+: seven lessons and about 77 minutes of study. The same subjects as v5, with less text and visual explanations. Also available on video.

OSWork Quick on video, with Nei

All seven lessons in one complete video per language, with Nei’s avatar and voice, illustrations and downloadable SRT subtitles. Choose a language to watch.

Your next conversation can pick up where the last one stopped.

OSWork turns isolated requests into organized work: define the objective, choose the context, complete a small task and check the result. Start with chat and progress to a restricted bot and supervised operation on a VPS.

What you build

  • A project folder with inputs, outputs and your own instructions.
  • A record of decisions and failures, alongside a reusable Skill.
  • A Git history for reviewing and recovering changes.
  • A query bot with fictional data and an operations plan.

The technical v2 path, module by module

Expand each module to see its six topics. Links open the full course in the language of this page.

Module 1 — Models: choose by task

Compare models with a real task and a quality criterion.

  1. AI as a work system
  2. Kinds of AI and what they are for
  3. Model is not reasoning effort
  4. Create a quality rubric
  5. Understand access and billing
  6. Define the autonomy
Open the full module
Module 2 — Chat, Work and Desktop

Write a work order with inputs, output, and review.

  1. Chat resolves a conversation
  2. Work receives an order
  3. Desktop brings the files closer
  4. Local and cloud are execution choices
  5. Write a delivery contract
  6. Review the received work
Open the full module
Module 3 — Terminal and Codex in practice

Open a training project in Codex and produce a verifiable change.

  1. Terminal is an entry point
  2. Install from the official source
  3. Authenticate without spreading secrets
  4. Enter the folder before requesting
  5. Do a first reading task
  6. Use permissions and recovery
Open the full module
Module 4 — Folders, Markdown and secrets

Build the digital house and separate knowledge from credentials.

  1. A folder represents a context
  2. Markdown is organized text
  3. Each file has a job
  4. Secrets are not shareable knowledge.
  5. Ignore before the first commit.
  6. Do a context cleanup.
Open the full module
Module 5 — AGENTS, Skills and memory

Create project instructions and a reusable capability with a review criterion.

  1. AGENTS.md guides execution
  2. Global and project complement each other
  3. Skills package procedures
  4. Memory needs maintenance
  5. Failures become small protections
  6. Test reusable capability
Open the full module
Module 6 — Git and GitHub without losing work

Save a version, inspect differences and recover a training change.

  1. Git is the project's history
  2. Observe before staging
  3. Save a version with intent
  4. Clone and update carefully
  5. Recover without deleting history
  6. Publish only what you reviewed
Open the full module
Module 7 — Telegram as work interface

Run a restricted query bot and understand where AI comes in.

  1. Telegram is the interface, not the agent
  2. Create the bot and protect the token
  3. Authorize people and actions
  4. Start with long polling
  5. Connect capabilities in stages
  6. Test operation and failures
Open the full module
Module 8 — VPS from scratch and 24/7 operation

Prepare a deployment plan, monitoring, backup and service verification.

  1. A VPS is a machine under your responsibility
  2. Enter via SSH and preserve access
  3. Install only what is necessary
  4. Protect the network and credentials
  5. Systemd supervises the process
  6. Availability requires a care routine
Open the full module

Study, take notes and continue later.

Mark topics as read, record questions, highlight passages and take notes. Adjust your reading preferences and export your journey. Progress stays in your browser and is separate for each language.

Your first workspace starts with real files.

The kit includes templates for AGENTS.md, memory, decisions, failure records, task contracts, a report Skill and a VPS plan. It includes a Python bot with an offline self-test.

The training bot is deterministic: it does not call an AI model or execute messages as commands. AI integration is explained as a later step.

Start with a small task.

  1. Choose the first module and compare two options available in your environment.
  2. Create a practice folder and adapt the task contract from the kit.
  3. Save the result, record a question and return to your journey.

Before you begin

Do I need programming skills?

You can start with the fundamentals without programming. Technical labs introduce the terminal, Git and Python step by step.

Does studying have a cost?

You can read the content without an account. Subscriptions, API usage and VPS hosting are separate services chosen by the learner for the labs.

Does the course create an unsupervised autonomous agent?

The course teaches limits, verification, logs and recovery. Running 24/7 requires configuration and monitoring; availability is not guaranteed.

Are the examples in my language?

Yes. Lessons, explanations, interface and kit are translated. Commands, file names and variables remain consistent across the three editions.