Before the session
Welcome: what OpenCode is and how this workshop works
We're glad you're here. This workshop is built for a true first time: it assumes nothing about AI agents, terminals, or programming. By the end you'll have an agent running on your own laptop and, above all, you'll understand what's happening when you use it.
When and where
- Wednesday, September 30, 2026, from 5:30 p.m. to 7:00 p.m., Mexico City time.
- Live, over video call. The link to join is on the workshop page, on the platform, while you're signed in.
- It's recorded. The recording is sent by email to everyone who saved a spot.
First things first: what OpenCode is
OpenCode is an open-source coding agent that runs in your terminal. Let's take it piece by piece:
- Agent: a program that doesn't just answer, it acts. It can read files, write files, and run commands on your computer.
- Open source: its code is public. Anyone can see how it's built, and using it costs nothing.
- Runs in your terminal: it's not a web page. You open it inside your computer's command window (the next lesson explains what that is).
An analogy: think of OpenCode as an assistant sitting at your desk. It can open your folders, read your papers, and write new documents. But the assistant doesn't think on its own: every time it has to decide something, it calls an expert on the phone who is somewhere else. That expert is the AI model, and the company that connects you to it is the provider.
| Piece | What it is | Where it lives | Analogy |
|---|---|---|---|
| OpenCode | The agent: reads, writes, and runs | On your laptop | The assistant at your desk |
| Provider | The company that gives access to models | On the internet (the cloud) | The phone company |
| Model | The AI that understands and answers text | On the provider's servers | The expert on the other end of the line |
| Your key | A password that identifies you | Stored on your laptop | Your customer number |
That's why OpenCode doesn't come with a model inside: you choose which provider to connect it to, and for that you need a key. We take care of it before the session.
The workshop map
This diagram sums up everything you'll put together. If you ever get lost, come back here.
YOUR LAPTOP THE CLOUD (internet)
+-----------------------------------+ +------------------------------+
| | | |
| You | | Model provider |
| (you type a request) | | (OpenRouter, Zen, Groq) |
| | | | | |
| v | request | v |
| OpenCode ----------------------------------> AI model |
| (the agent) <------------------------------- (thinks and answers) |
| | | answer | |
| v | +------------------------------+
| agent-workshop folder |
| (reads and writes files here) |
| |
+-----------------------------------+
THE LIVE SESSION, IN 5 PARTS
+------------+ +------------+ +------------+ +------------+ +------------+
| 1. Agent |-->| 2. Install |-->| 3. First |-->| 4. First |-->| 5. What not|
| vs chat | | and connect| | job: it | | change: it| | to ask yet|
| | | OpenCode | | reads | | writes, | | and where |
| | | | | | | you review| | to go next|
+------------+ +------------+ +------------+ +------------+ +------------+Notice two things. First: the model never touches your files. OpenCode does the reading and writing, on your machine; the model only receives text and answers with text. Second: when OpenCode reads a file, it sends its content to the model along with your request, so the model can understand it. That's why this workshop uses sample data.
What we'll do live
In 90 minutes, hands on the keyboard:
- Understand what an agent is and how it differs from a chat.
- Install OpenCode and connect it to a model.
- Give it its first job: read a folder and explain what's in it.
- Ask for its first change, and review it before accepting it.
- See what not to ask it yet, and where to go next.
What you'll understand at the end (not just do)
Doing the steps is half of it. The other half is knowing why they work, so you can repeat them on your own tomorrow. By the end you'll be able to explain in your own words:
- What a terminal is and why agents live there.
- What a provider, a model, and a key are, and why the key is secret.
- Why an agent makes several requests for a single job, and what that has to do with free plan limits.
- Why the agent asks your permission before changing something, and how to review what it proposes.
- Why you shouldn't trust it blindly, and how to check one fact quickly.
How the three modules fit together
+-------------------+ +-------------------+ +-------------------+
| 1. BEFORE | | 2. LIVE | | 3. AFTER |
| | | | | |
| You prepare your | ---> | You install, | ---> | You practice on |
| laptop and get | | connect, and | | your own and see |
| your key | | give the agent | | where to go next |
| | | jobs | | |
+-------------------+ +-------------------+ +-------------------+
~15 minutes 90 minutes 10 min a day| Stage | What's there | Why it goes there | When |
|---|---|---|---|
| Before | Prepare your laptop and get a key from a model provider | These steps don't depend on anyone and eat up time live | Before Wednesday |
| Live | A single lesson with the five parts of the session, with commands and requests ready to copy | That's where it helps to have support if something fails | During the session |
| After | Where to go next and how to get your certificate | What you don't repeat, you forget | In the following days |
During the session, keep this page open in another tab. That way you copy the commands instead of retyping them from the shared screen: an extra space or a wrong letter in a command is the most common first-time mistake.
The only thing we ask beforehand
Do the next two lessons. They take about 15 minutes:
- Prepare your laptop: download Warp or Visual Studio Code, open a terminal, and create the practice folder.
- Get your model key: create a free account with a provider and save your key.
If you arrive without doing them, that's fine: the first minutes of the session are for exactly that, and we don't move on until the room is ready. But getting the key is the slowest thing to do live, so if you only have five minutes, start there.
Check your understanding
1. If OpenCode is the agent, why do I also need a provider?
Because OpenCode doesn't come with a model inside. OpenCode is what reads and writes on your laptop, but to "think" it asks an AI model running on a provider's servers. Without a provider (and without a key), the assistant has no one to call.
2. Where do your practice files live, on your laptop or in the cloud?
On your laptop, inside the agent-workshop folder. OpenCode runs on your laptop too. The only thing in the cloud is the model, and it receives the text OpenCode sends: your request and the content of any files it read.
3. Why is it worth getting the key before the session?
Because it's the slowest step: creating an account on a website, confirming your email, and generating the key. The rest (installing and connecting) takes minutes during the session.
Summary
- OpenCode is an open-source agent that runs in your terminal: it reads, writes, and runs things on your laptop, and to think it connects to a model through a provider you choose.
- The workshop is Wednesday, September 30, from 5:30 p.m. to 7:00 p.m. Mexico City time, and it's recorded. The recording arrives by email.
- This material has three parts: before, live, and after. Before the session: prepare your laptop and get your key. Next lesson.