Module 0

Orientation & Setup

Orientation & Setup

Welcome! You are starting a program to learn agentic coding using the SOLO Protocol.

What is the SOLO Protocol? The SOLO (Single Operator Loop Orchestration) Protocol is an engineering methodology designed to transition developer workflows from ad-hoc "vibe coding" into disciplined, specification-driven software construction. To understand the core philosophy, history, and mechanics of the protocol before you write your first line of code, read the original blog article: The SOLO Protocol.

How This Program Works

You'll build one project across the entire program — a command-line study tracker in Python. No web frameworks. No databases. No frontend. Just Python, files, and your terminal.

Each module teaches one idea from the SOLO protocol. They build on each other, so do them in order. Every module has:

  1. What you need to know first — background concepts explained simply
  2. The SOLO idea — what the pillar is and why it matters
  3. Lab — hands-on work where you apply the idea to your project
  4. Checkpoint — something you produce to prove you learned it

A note on cost: GitHub Copilot is free for verified students through GitHub Education — sign up with your *.edu email. This gives you access to Copilot Chat in VS Code and OpenAI Codex. The program is designed to work within these free-tier limits if you're deliberate about your sessions. Each module estimates how many AI conversations you'll need. Don't burn through your usage aimlessly — that's literally one of the things SOLO teaches you to avoid.

Setup (Do This First)

1. Install VS Code and GitHub Copilot

  1. Download VS Code if you don't have it
  2. Open VS Code, go to the Extensions panel (⌘+Shift+X on Mac, Ctrl+Shift+X on Windows)
  3. Search for "GitHub Copilot" and install it (it's by GitHub — the verified one)
  4. Sign in with your GitHub account when prompted
  5. Make sure you have a Copilot subscription (free for students — see GitHub Education)

How to use Copilot Chat: Press ⌘+I (Mac) or Ctrl+I (Windows) to open the chat panel. This is where you'll type instructions to the AI. You can also click the Copilot icon in the sidebar.

In Copilot Chat, enable Agent Mode — it lets the AI edit files, run commands, and iterate on code autonomously (instead of just suggesting snippets). Look for the mode selector at the top of the chat panel.

2. Install OpenAI Codex

Codex is a terminal-based agentic coding tool. While Copilot Chat works inside VS Code, Codex works in your regular terminal — you describe what you want in plain English and it writes and runs code for you.

  1. Install via npm: npm install -g @openai/codex
  2. If you don't have npm, install Node.js first (which includes npm)
  3. Run codex --help to verify it works
  4. Codex uses your GitHub Copilot credentials, so if you're signed into Copilot, you're good

When to use which:

  • Copilot Chat (in VS Code): Best for working with files you have open, editing code, and conversations where you want to see changes in context
  • Codex (in terminal): Best for autonomous tasks like "implement this whole module" or "run all tests and fix failures" — it works more independently

You'll use both throughout this program. The instructions will tell you which one to use for each step.

3. Install the SOLO workflow files

These are text files that teach the AI how to behave when you give it specific instructions. They're like recipes — you point the AI at a recipe file and say "follow this."

  1. Open your terminal
  2. Navigate to where you want your project:
    • Mac/Linux/PowerShell: cd ~/Projects (or wherever)
    • Windows Command Prompt: cd %USERPROFILE%\Projects
  3. Create your project folder: mkdir study-tracker && cd study-tracker
  4. Clone and set up the SOLO workflows:
    • Mac/Linux: 
      git clone https://github.com/mario-guerra/solo-protocol-release.git .solo-temp
mkdir -p workflows
cp .solo-temp/workflows/*.md workflows/
rm -rf .solo-temp
    • Windows Command Prompt: 
      git clone https://github.com/mario-guerra/solo-protocol-release.git .solo-temp
mkdir workflows
copy .solo-temp\workflows\*.md workflows\
rmdir /s /q .solo-temp
    • Windows PowerShell: 
      git clone https://github.com/mario-guerra/solo-protocol-release.git .solo-temp
mkdir -Force workflows
copy-item .solo-temp/workflows/*.md workflows/
remove-item -Recurse -Force .solo-temp
  5. Open the study-tracker folder in VS Code: code .
  6. Verify: you should see a workflows/ folder in your Explorer panel containing files like architect.md, critique.md, code.md, etc.

How to use SOLO workflows in Copilot Chat

Some agentic IDEs (like Cursor or Antigravity) let you type /architect as a slash command. Copilot Chat doesn't work that way. Instead, you reference the workflow file and tell the AI to follow it. Here's the pattern:

Follow the instructions in #file:workflows/architect.md to architect my project.
Here are my requirements: [your requirements here]

The #file: syntax tells Copilot Chat to read that file and include it as context. The AI then follows the instructions in the workflow file.

Throughout this program, when you see an instruction like "run /architect", translate it to:

In Copilot Chat: Follow the instructions in #file:workflows/architect.md — [your request]

In Codex: Follow the instructions in workflows/architect.md — [your request]

Here's the quick translation for every SOLO command:

SOLO Command What to type in Copilot Chat
/architect Follow #file:workflows/architect.md — [your request]
/critique Follow #file:workflows/critique.md — critique the plan above
/revise Follow #file:workflows/revise.md — address these findings: [list]
/code Follow #file:workflows/code.md — implement based on the plan
/review Follow #file:workflows/review.md — review the code
/fix Follow #file:workflows/fix.md — fix the issues from the review
/test Follow #file:workflows/test.md — run all tests and fix failures
/sherlock Follow #file:workflows/sherlock.md — [describe the symptom]
/memory Follow #file:workflows/memory.md — save session context
/resume Follow #file:workflows/resume.md — restore context from .tmp/SESSION_MEMORY.md

You'll get used to this quickly. After a few modules it'll be muscle memory.

4. Verify Python

Run the version check in your terminal (you need Python 3.10 or higher):

  • Mac/Linux: python3 --version
  • Windows: python --version If it's not installed or is an older version, go to python.org and install it.
Back To

Program Overview

Next Module

Learn Git (The Basics)