Big Idea

The argument in one line.

OpenCode eliminates the attention bottleneck in AI-assisted coding by running multiple AI agents in parallel on different models simultaneously, letting you orchestrate rather than babysit.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

A solo founder or small team builder running multiple AI coding tasks sequentially and losing time switching between terminals and token budgets.
A developer who subscribes to multiple AI providers (OpenAI, Claude, Gemini) and wants visibility into usage across all of them simultaneously.
Someone shipping solo projects who needs to parallelize coding work but lacks a team to divide tasks across.

SKIP IF…

You're already using a dedicated CI/CD pipeline or team infrastructure to parallelize coding tasks — this targets individual workflows.
You only work with one AI provider or one coding task at a time and don't have switching overhead to solve.

TL;DR

The full version, fast.

OpenCode solves the core bottleneck in AI-assisted coding: attention. Instead of one terminal, one task, one model at a time, it lets you run multiple parallel sessions across different AI providers — Gemini for planning, GPT or Codex for execution — with usage tracking and completion alerts so your job shifts from babysitting terminals to intervening when needed. Setup covers four install paths (desktop app, IDE extension, CLI, or web), plus connecting provider API keys through the model selector. Key features include /init to generate an agents.md file that reduces token consumption on every future session, /review for security and code quality checks, custom skills for repeatable workflows, and custom slash commands like git commit and push. Context7 MCP adds library documentation lookup at a fraction of the cost of reading files directly.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →

Chapters

Where the time goes.

00:00 – 01:41

01 · Hook + Keith intro

Names the bottleneck (attention), introduces OpenCode and Keith background.

01:41 – 03:19

02 · Old workflow vs OpenCode

Whiteboard diagram: before = chaos of 3 terminals; after = OpenCode task list with usage remaining.

03:19 – 05:19

03 · What is OpenCode + agenda

Task orchestration for AI coding. Agenda: install, connect providers, build/plan modes, MCPs, skills, live demo.

05:19 – 09:55

04 · Install: Desktop, Extension, CLI, Web

Desktop app, Cursor/Windsurf extension, curl one-liner for terminal, opencode --web for remote browser access.

09:55 – 10:15

05 · Settings and appearance

Dark mode, Dracula/Tokyo Night theme, sound effects on task complete, shortcuts.

10:15 – 15:00

06 · Build vs Plan mode + model strategy

Plan with Gemini 3 Pro (high thinking), execute with GPT-5.2 Codex; token budget panel; split features into separate sessions.

15:00 – 18:02

07 · Connecting AI providers

Desktop: /model > Connect Provider. Terminal: /select. GLM 4.7 and MiniMax rank above GPT-5.2 Flash on benchmarks.

18:02 – 19:00

08 · Live demo: parallel sessions on health app

Bug fix via screenshot + Gemini planning + UI/UX improvement — three tasks simultaneously. Gamification (streaks) built live.

16:57 – 19:00

09 · Advanced: agents.md + /review

/init reads codebase and writes agents.md for token efficiency. /review catches bugs inline.

19:00 – 21:00

10 · Advanced: MCP servers (Context7)

Context7 MCP via global opencode.json. Add rule to agents.md so model calls Context7 automatically.

21:00 – 24:10

11 · Terminal CLI: split panes + model picker

Control-T = token cap, Tab = plan/build, Control-P = model. Terminal exposes more models than desktop.

24:10 – 25:57

12 · Custom skills and commands

Skills = reusable instruction sets in skills.md. Commands = slash shortcuts. Added git commit+push live.

25:57 – 28:03

13 · Mobile: OpenCode on iPhone via Termius

SSH remote login on Mac, Termius app, local WiFi. Cloudflare tunnel for remote access is a future video.

28:03 – 28:24

14 · Outro + community CTA

Like/subscribe, free AI community at rumjahn.substack.com, vibe coding course.

Atomic Insights

Lines worth screenshotting.

Running multiple AI coding agents in parallel across different models shifts your role from watching AI work to intervening only when needed — the bottleneck in AI-assisted development is not the AI's speed but the human's attention.
A model strategy that uses Gemini for planning and GPT-5.2 for execution — because Gemini plans more intelligently and the execution model doesn't need to re-derive the architecture — is a concrete application of right-tool-for-the-task thinking in a multi-agent workflow.
Splitting three related features into three separate sessions instead of one continuous chat cuts token consumption because each new session doesn't carry the dead weight of the prior session's context history.
The agents.md initialization command reads your codebase and generates a file that captures your project's conventions — reducing the tokens required for every subsequent session because the model doesn't re-infer what it already learned.
Context7 as an MCP that reads library documentation on demand — rather than dumping full API docs into the context window — is the token optimization that makes working with large external libraries sustainable at scale.
Adding 'always use Context7 when I need library documentation' to the agents.md rules file converts a manual invocation into a default behavior — the model does it without being asked on every subsequent session.
Running OpenCode via SSH on a mobile phone (Termius → same-network connection) converts a laptop-bound coding environment into one you can monitor and direct from anywhere on the same network.
Token usage tracking that shows how much credit remains across multiple subscriptions solves the problem of unknowingly running out mid-session and having to switch providers without realizing the context was lost.
The /review command that scans the codebase for security vulnerabilities, unused code, and styling issues before a major feature addition is the quality gate that prevents accumulated technical debt from becoming a refactoring emergency.
Attaching a screenshot to a bug report and saying 'the time frame word is out of place' gives the model visual context that a text description alone cannot provide — which is why multi-modal input changes the quality of bug fix instructions.
The terminal version of OpenCode exposes more model options and more slash commands than the desktop app, which is why developers who prioritize capability over UI convenience prefer the TUI over the graphical interface.
A git commit and push command built as a custom OpenCode command converts a multi-step terminal ritual into a single slash-key invocation — the same pattern that makes any repetitive terminal workflow worth packaging as a command.
Running gamification planning (Gemini), bug fixing (GPT-5.2), and UI improvement brainstorming in three simultaneous sessions on the same project is the parallel workflow that makes AI coding a genuine throughput multiplier.
The agents.md file is to OpenCode what CLAUDE.md is to Claude Code: a persistent behavioral contract that accumulates project-specific knowledge and reduces the cost of context-loading on every subsequent session.
Warp terminal as an AI agent for the terminal itself — used to find config file locations, add code snippets, and troubleshoot installation errors — is the meta-tool that eliminates the friction of navigating developer tooling as a non-expert.

Takeaway

Steal the parallel-agent frame.

JoeFlow + builder playbook

The bottleneck isn't AI speed — it's your attention. The moment you frame your tool as a supervision cockpit instead of a chat window, everything changes.

Lead with the bottleneck not the feature: 'your attention is the limit' lands harder than 'run multiple sessions.'
The agents.md / CLAUDE.md pattern is the same insight — front-load context once, save tokens on every future task.
Token budget awareness as a workflow decision is a strong content angle for the $20 stack audience.
The Sequential vs Parallel whiteboard diagram is a clean visual for a short — one frame, one thesis, steal the format.
His Claude API warning is worth a JoeFlow response video — clarify what JoeFlow actually does differently.
The Termius mobile SSH demo is an easy hook for a short: 'I coded from my iPhone on the bus.'

Glossary

Terms worth knowing.

OpenCode: An open-source AI coding tool that orchestrates multiple AI agents in parallel sessions, letting you run different models on different tasks simultaneously instead of working one terminal at a time.
Token: The unit of text AI models read and generate, used by providers to meter usage and billing. Longer prompts, files, and responses all consume tokens against your plan's limit.
Terminal (CLI): A text-based interface where you type commands instead of clicking buttons. Command-line tools often run faster and offer more options than their graphical desktop counterparts.
TUI: Text User Interface — a keyboard-driven app that runs inside a terminal but uses panels, menus, and shortcuts instead of raw command output.
IDE: Integrated Development Environment — a code editor like Cursor, Windsurf, or VS Code that bundles writing, running, and debugging tools into one app.
Extension: An add-on that plugs new functionality into an existing IDE so a tool like OpenCode runs inside the editor you already use.
curl: A command-line utility for fetching files or running install scripts from a URL. It's commonly used to detect your operating system and pull the right installer automatically.
Warp: An AI-powered terminal application that can interpret natural-language requests, run shell commands, and troubleshoot errors for you inside the command line.
Ghostty: A fast, GPU-accelerated terminal emulator used as an alternative to the default macOS Terminal app.
Plan mode vs. build mode: Two operating modes in AI coding tools: plan mode reasons through approach and architecture without changing files, while build mode actually writes and edits code.
Thinking effort: A setting that controls how many tokens a reasoning model is allowed to spend deliberating before answering. Higher effort yields deeper reasoning at greater cost.
Context window: The maximum amount of text a model can hold in active memory at once. Longer chats and bigger files eat into this budget and force older content to be dropped.
Session: A single isolated conversation thread with an AI agent. Splitting work across sessions keeps each thread short, preserves context, and reduces token spend.
API key: A secret string that authenticates your account when an app calls a provider's API. You generate one in the provider's dashboard and paste it into the tool to unlock paid model access.
GLM 4.7: An open-source large language model from Zhipu AI, often offered free inside coding tools and competitive with leading paid models on common tasks.
MiniMax: An open-source large language model from a Chinese AI lab, used as a low-cost alternative to OpenAI, Anthropic, and Google models.
GPT-5.2 Codex: A code-specialized variant of OpenAI's GPT-5.2 tuned for software engineering tasks like refactoring, debugging, and multi-file edits.
Gemini 3 Pro: Google's flagship reasoning model in the Gemini family, used here for high-level planning and architectural thinking before code is written.
MCP: Model Context Protocol — an open standard that lets AI agents connect to external tools, data sources, and services through a uniform interface.
Playwright MCP: An MCP server that gives an AI agent the ability to drive a real browser — navigating pages, clicking, and filling forms — to test or automate web tasks.

Resources

Things they pointed at.

03:13productopencode.ai ↗

05:58toolWarp terminal ↗

19:23toolContext7 MCP ↗

25:57toolTermius SSH app ↗

28:20linkKeith AI community (Substack) ↗

Quotables

Lines you could clip.

02:39

“My job now isn't to watch AI work. My job now is to intervene when it's only needed.”

Tight two-sentence thesis; no setup needed; reframes the job description of a builder→ TikTok hook↗ Tweet quote

09:22

“I'd trade money for time any day of the week.”

Clean one-liner on paid vs free model philosophy→ IG reel cold open↗ Tweet quote

10:05

“Instead of making it one long continuous chat one after another, which consumes a lot of tokens, I split them up into new sessions each one so that it saves me money.”

Actionable tip with clear before/after; easy to demo→ TikTok hook↗ Tweet quote

The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

metaphoranalogystory

00:00Most AI coding tools still work like this. One task, one terminal, one conversation at a time.

00:06What if I told you you can code much faster because of one bottleneck, and that bottleneck is your attention.

00:13Today, I'm gonna show you how I use OpenCode to run multiple AI tasks in parallel, track usage, and stop babysitting terminals.

00:22This is a step by step tutorial and how I actually use OpenCode, a real workflow under real constraints to ship faster without a team.

00:32I'm Keith. I built and exited companies, taken one public, and now I help solopreneurs and builders ship faster by turning AI into real systems. I've used pretty much all of the coding tools, Cursor, OpenAI Codecs, Cloud Code.

00:47You name it. I've tried it. They're all good at generating code.

00:49None of them solved the thing that was slowing me down, running multiple tasks and knowing what's happening. So let me show you my old workflow. I'm subscribed to Gemini, Claude, and OpenAI, and I used to open a terminal for each one.

01:03But the problem was when I try to run multiple tasks in three different windows, I don't know which one has failed, which one is done, and most importantly, I don't know how much credit I have left for each one. So multitasking was a complete nightmare.

01:16What really ended up happening was I just had one window and I kept using it until the credits ran out, and then I switched to a different one. So I can only do one task at a time. After I switched to OpenCode, it really changed my world because now I can run multiple tasks in parallel and will tell me how much usage I've used.

01:33It'll alert me when a task is done, and I can run multiple models on the same project at the same time. Absolute game changer. Quick context, what is OpenCode?

01:42Here you see OpenCode is open, and I think of it basically as a task orchestration for AI coding instead of just one terminal and one task at a time.

01:52Here you see that I have multiple sessions going on, and for each one, I can choose a different model. So I can choose to run GPT 5.2 codecs for one task. Let it run, and I can run another session using Gemini three Pro.

02:08Let that task run as well. And when it's done, it'll alert me so I can run four or five different tasks at the same time, and that's what OpenCode does. I'm gonna go through the step by step instructions from installation all the way to very advanced features with OpenCode, but it's important to understand some high level concepts.

02:26So why am I excited about OpenCode? It's because previously I was doing sequential one task at a time, and now I can do parallel task and run multiple agents coding at the same time.

02:37So my job now isn't to watch AI work. My job now is to intervene when it's only needed.

02:43So what are we covering in this video today? We're gonna be covering from installing OpenCode as a terminal or downloading the desktop app and even showing you how the web version works to connecting to all your different AI providers like Gemini, OpenAI, how to choose the free models and your paid models, choosing between build and planning, adding MCP servers, adding agent skills, and how to integrate it with tools like warp, and finally, giving you a live demo of how I use it in my workflows, working on a project that's generating revenue and running multiple tasks.

03:19So the first thing you need to do is go to opencode.ai. And the first option I'm gonna show you and the easiest one is to download the desktop app. So you can click on download now, and I'm using a Mac app, so I'm gonna download Mac OS silicon.

03:34Alright. It's downloaded. I double click on it and just follow the instructions, and you should be granted with a desktop app like this.

03:42Alright. The next option is that you can choose extensions. And if you're already using Cursor, Windsurf, or any of the coding IDEs, you just press install, and it'll get installed into those IDEs as well.

03:53Very simple. And then you have the option of installing it through the terminal. Now what's the difference between installing through an extension or downloading the desktop app?

04:02Extensions like Cursor give you basically an app with an interface, so it makes it easier to have visibility, run multiple tasks, gives you alert, gives you gives you a lot more information to work with.

04:15But some people prefer terminal because it's very focused, and if you're familiar with terminal commands, it can be faster. I actually like terminal because I found using terminal, for some reason, seems to be faster than using an IDE, and you have all these options.

04:32Right? Curl, n p m, bun, brew, power. The most straightforward one is actually using this command, the curl command, because it will identify your operating system, and then it will choose the best option to install on your machine.

04:46Because if you're using Linux, you might use paru. If you're using macOS, you might use brew install. But if you're unsure, just run the curl command.

04:54So open your terminal. So you can go to terminal and open your terminal, but I use this different tool called ghosty.

05:02Same thing. Just type this in, enter, and it will do its magic. So those are the three options you can use to install OpenCode, and there's a hidden option which is you can run it on the web.

05:14So after you've installed it in the terminal, you just run this command, open code web, copy, and now it runs in my browser.

05:24So you can see that I can run open code there. Okay. So why would you wanna run a web version of open code?

05:30I think the most useful use case is where you want to run OpenCode on one machine, on one particular machine, and you wanna access it using a different machine, maybe your mobile phone. And so you can connect to it, open a browser, and start coding away while the source code of your project is on the computer.

05:49A bit of pro tip on installing the CLI, command line interface, in the terminal is that sometimes you run into trouble.

05:56No worries. You can download a tool called warp dot dev, and what warp dot dev is is an AI agent for the terminal.

06:04So you can just copy and paste your command into warp dot dev.

06:09And if there's any issues, you can just ask AI agent to fix it for you, and it will fix all the issues for you and make sure you get it installed. And once you have installed, all you need to do is type in open code and voila, you're inside the terminal and open code is running. The first thing I wanna show you inside the open code app is the settings because everyone loves dark mode.

06:32So you can come here, go to appearance, and go ahead and set it to dark. You can choose the font.

06:38You can also choose a theme, Dracula, Tokyo night. So pick whatever you want, and you have the option to turn on agents, permissions, errors, and you can also have a sound effect when your task completes, and you can also add in shortcuts.

06:54So that's basic settings. Now I'm gonna go through some of the features that OpenCode has. Right here, there's three different options you can choose from.

07:02There's build and there's plan. And how I use it is that when I'm thinking of adding a new feature, I usually go into plan mode first before I build it. So then it thinks through the logic and then it builds.

07:15Then you can choose your model. Now it comes with a lot of free models. GLM 4.7 is actually pretty good, and you can do a lot with free models already.

07:24But in the end, I think paid models are better, and I'd trade money for time any day of the week. Here, I've already connected Google, OpenAI, and Anthropic.

07:35And usually when I'm planning, I would choose Claude, but recently Claude announced they don't support third parties, so I do not recommend using Claude with open code because you might get banned.

07:48So when I'm planning, Gemini is better at planning, and so I'm gonna use the most powerful one, which is Gemini three pro preview. And then thinking effort, I can click, and I wanna set it on high because I wanna think carefully.

08:02And then when I'm ready with the plan, I click on build where it will actually start coding, and I would use either 5.2 or 5.2 codex. If it's more advanced feature, then I would use codex.

08:15But if it's something simple like adding a button or just fixing a bug, I would choose 5.2. And if that fails, then I would use a higher context one. So being mindful of my token use.

08:26And thinking effort is basically how many tokens you wanna cap your request at. This is an interesting button and actually is a very important tool because it shows how much usage you have left. If you're like me and you're on a $20 plan, things run out pretty quickly, but if you're a $100 plan, it might take you a while to run out, but it's good to see how much context you have left.

08:48And another thing I do is sometimes I attach screenshots saying that some certain UI elements are out of place, and I would just screenshot it and attach that and say fix this or show the AI what's being made. So those are the options here.

09:02I'm gonna take a second here to explain my model strategy because I'm subscribed to Claude, OpenAI, and Gemini on a $20 plan. And I like OpenCode because it forces you to think about using which AI and which model is most appropriate for your task.

09:19Right? And I basically use Gemini for high level planning, and I use OpenAI once the planning is done to execute the code.

09:30And depending on the task, if it's complex, I'll use GPT 5.2 codec. If it's just, uh, like, a simple bug fix or moving a button, I'll use a lower model, and I'll set the token limit to a lower rate.

09:41So this is, like, representing the token usage. So this is a good thing to just get in the habit of choosing the right model and thinking about how many tokens you need for what task.

09:52So another technique I use to save on my precious tokens is that, let's say I have three features. Instead of making it one long continuous chat one after another, which consumes a lot of tokens, I split them up into new sessions each one so that it saves me money and contacts as well.

10:12So a little trick that I use to save tokens. OpenCode comes with all these free options, but if you wanna be like me and connect to Anthropic, OpenAI, and Google, you need to connect to your provider.

10:25I'm gonna show you how to connect to different AI models in the OpenCode desktop app. You click slash, and then you go to model.

10:35And then on the top right, you can click on connect provider, and then you just choose the provider that you want. Let's say Google, and then you need to enter your API key.

10:45I'm also gonna show you how you can add models if you're using the TUI terminal app, and then I'll go into how you can actually get the API key easily with the help of AI agent without spending, like, ten minutes clicking around. Here, I've opened my terminal, and here I can type in slash, go down, and choose select. And then you can see that I've connected to Anthropic, OpenAI, and Google, but there's so many options here.

11:11This is the only tool where I've seen you can basically connect to everything that is out there. And I wouldn't underestimate the Chinese AIs because right now, in sixth place and seventh place, you have GLM 4.7, which is open source, and MiniMax, which is also open source.

11:30That's beating ChateaPT 5.2 and Gemini three Flash. So you can get pretty good results with these other options as well, and it's much cheaper to subscribe to them.

11:41So let's click on one of them. I have Perplexity, so let's click on Perplexity, and then you enter your API key.

11:48And how do you do that? Well, I use a browser called Perplexity Comment, and what I can do is I can come here on the top right assistant, and the AI agent will actually click around the interface and find the API key for me.

12:02So saving you time from clicking and trying to find all these API keys for Google or ChatGPT. Just quick note that you never wanna pass your API keys or passwords into AI. And so this hack to do this is I need to create an API key for OpenCode from Perplexity.

12:20Can you set up the API key so that it goes up to the stage where the key is ready for me to manually copy and paste? Please do not copy and paste it for me. So right now, I've given instructions to basically set up the API key, but don't copy it.

12:35Stop there, and then I'll manually copy it and paste it into OpenCode. And now it should save you a ton of time in clicking around on interface. Okay.

12:43Now it's done. I'm gonna copy and paste. Okay.

12:47And now it's connected, and I can choose which one. And then now I should be able to go in Perplexity.

12:54I can choose Sonar, Sonar Pro, reasoning. So it's connected. So that's how you connect your different AI providers to OpenCode.

13:02Now I'm gonna actually show you how I actually use OpenCode. When you first come in, you're gonna open a project.

13:09If you're starting a new project, then you gotta create a new folder. And so I'm gonna open a project, and here I've opened my project. To give you some context, I have an app here, which is a health data analyzer app where you sync your health data and it gives you analysis, and you can do AI analysis of your health data as well.

13:29So here you can see that I've got an issue here with this not displaying correctly. So let's take a screenshot, and we'll use that in a bit.

13:38And now that I've connected, I wanna first fix that bug. I'm going to attach the screenshot and say for the graphs, the time frame word is out of place.

13:52Can you fix it? And then I can click send.

13:55So it's using the attach image function, and I've chosen my model.

14:03While that is running, this will display and it's doing its magic. I'm gonna start a new session. I'm gonna plan, and I'm gonna choose Gemini, which is different model, and say, for my health analyzer app, I wanna add some gamification ideas to the products so it makes it more sticky, maybe adding streaks or something.

14:25But the end idea is to make sure people sync their data on a regular basis and kinda reward them for the when they actually do sync their data.

14:35So I'm gonna let it plan, hit enter, and you can see here that this is running. And, also, my other session is also running.

14:44So this is where multiple thing can run at the same time, I and can start another session and say, can you some ideas on how to improve the UI, UX of the product? I can run that too. And you'll see that if we check-in, it it's showing you the progress and the time it's elapsed and also how much code is being changed with the plus sign and or how much is being deleted.

15:07You can see all the changes it's making in the code base. And when you see this blue icon, it means that the task has been completed.

15:17You can also turn on notification where it will show up on the top right of your Mac that a task has been completed. I turned it off because it was just popping up all the time, but if that's something you like, you can have it run as well.

15:30And now all three tasks are done, and let's see if it actually fixed the problem. So you see this is out of place.

15:37Let's turn it off and run it again. Alright. And the time frame wording is gone completely, and that's been fixed.

15:46So happy with that. And for the gamification, it's giving me some suggestions on adding basically a badge and milestones, and I kinda like the idea.

15:58So let's try to build it. So I turn on the build mode right now. I'll switch back to GPT 5.2.

16:05Yes. Yes. Proceed with this plan, and then let's see what happens.

16:08So you can run multiple sessions at the same time, but let me show you something else. I don't have just one product, but I have multiple products.

16:18I can open another project, and in there, I can also run multiple sessions. So I can be working on two projects at the same time, and each project can be running multiple sessions at the same time as well.

16:30So you can be running many, many things at the same time. Alright. And it's added the streaks, and so let's go back into the app, and let's check it out.

16:39And there it is. I've added a street count, three days, seven day, fourteen day, thirty day.

16:45Looking really good. I think I could do a little bit more by showing more than just three, seven, fourteen, and thirty, but no, it's done a good job.

16:56So looking very good. Okay. Now let's look into some of the more advanced features it offers and one of the secrets is that you can type the slash command and a lot of cool things show up.

17:08Okay. Let's start with the first command, is the init instruction. It creates an agents dot m d file.

17:14I would highly recommend doing this because OpenCode will read all your files and your folder structures and create a file that makes it more efficient to run-in the future. So if you wanna save your tokens and do more faster, definitely worth running this command.

17:28And now it's running, and you can see that it's running some code style guidelines, formatting, naming conventions, error handling, and creating this agents dot MD file.

17:40And just to showcase this parallel feature, I'm gonna start a new session, and let's do the second thing, which is review. What does a review command do?

17:50It basically looks at your code, checks for security vulnerabilities, any potential bugs or issues or styling issues, And it just goes through your code and make sure it's in a good state.

18:01And I run this from time to time and I also refactor my code from time to time if I'm adding a whole bunch of features. Because sometimes you might add in some code that's not used anymore and you wanna remove that to keep your file size smaller.

18:14And here it's come back with some issues and that's great.

18:18You know, it's realized there's a issue in my current streak and, uh, it should fix that. I think it's a good idea to fix it. Can you please fix it?

18:27And then here's the agents dot m d, and let's take a look. It's realized as an Xcode project, and it's just all these things, naming conventions, concurrency error handling.

18:40And so this this is actually really useful because in the future, it will reduce the tokens required to run my project. So useful to run and always initialize a new project. Okay.

18:51Next thing we're gonna do is look into MCPs. So you can toggle on and off your MCPs, and here I've added Playwright.

19:00It has ability to open a browser and pretend to be user and click around and do stuff for me using an AI agent. The MCP I want to add is Context seven and basically is an MCP that will help me save tokens again by allowing it to read API documentation and libraries at a much lower cost than reading everything at once.

19:19So to use context seven, I need to sign up and generate an API key, and then I need to add this to the configuration file.

19:29So if I come to open code, it says you have to add something like this to my configuration file. And now I have a hack here where I can use warp dot dev that I mentioned earlier, and I wanna say find the global open code dot JSON.

19:45That's where the configuration file is because I hate clicking around and finding the right directory and the right file, and this allows me to kind of quickly find the global file. Once it's found a global file, I'm going to ask it to add this bit of code inside, and I will share the code in the description.

20:05You can get it there as well. I'm adding the contact seven MCP to the open code file, and then I accept the change. Can you open the file using Sublime note editor so I can add in the API key?

20:22So here you see that it's added this to my config file, and all I need to do is copy and paste my API key. I signed up for contact seven, and I get, like, a free API key with a thousand requests.

20:35So I'm gonna generate that and put it in. I'm gonna label it open code.

20:41Paste it in. Save. And then let's close and restart open code and see if it's added successfully.

20:48So I'll come here, MCP, toggle MCPs. There you have it.

20:527 is added. Even though I added 7, I keep forgetting to call 7.

20:57You have to say use 7 specifically. So inside the documentation, actually, suggests adding a rule.

21:05Always use 7 when I need library documentation. So how do you add this rule to OpenCode? So remember the agents dot m d file when we initialize the project?

21:16Well, we're gonna go do that. I'm gonna go to OpenCode and say, this rule to the agents dot d.

21:28And there you go. It's added the rule to the agents dot m d file, and from now on, it's gonna save me tokens and call contact 7 on its own without me calling it. And so that's how you add MCP servers to improve your coding experience combined with rules.

21:44Now that we've gone through the desktop app, let's go through the terminal app and check out the command line interface. I'm gonna use warp here, but you don't have to use warp.

21:53You can just open terminal or you can use ghosty or whatever terminal app you use. And the first thing I need to do is to open my directory.

22:02So I go to CD, go to projects, and then I drag and drop the location. This means change directory, and then it'll ask you to optimize and, of course, I do wanna optimize.

22:15Okay. Now how do you bring up OpenCode? All you need to do is type in OpenCode and voila.

22:20You're in. So you can do the same things you can do on your desktop app, but now you need to use these keys. So control t allows me to control the variant.

22:31So how many tokens kept? I can press tab to change agent, so planning or building, and then control p allows me to switch models. So I can choose and you can see that the desktop version allows me to have more selection.

22:46The desktop only gives me three options, but the terminal is way more powerful, Allows me to choose way more options than the desktop app. Super powerful stuff here.

22:55So that's why you would use the terminal because it's just got way more functionality. Also, I type the slash button, I have more options to choose from.

23:05So I can choose a theme, so let's say Tokyo Night Solarized, I can go orange.

23:13I'm gonna stick with the original because I like it, but you can change the color scheme. I can see the status and I can go to sessions and I can see all the sessions are happening right now. Right now, I can only run one task at a time.

23:26What if I wanna be like the desktop and run multiple sessions at the same time?

23:31Well, what I can do is I can go to split pane right and I can keep clicking it and open three or four panels and then I do open code, open code, open code, and all of a sudden I have four tabs open and I can run four sessions at the same time in the same project.

23:51What if I wanna run another project? Well, I can just click here and open a new tab and I can open another project and I could have multiple sessions within that project as well.

24:04So you can do the same thing you did in desktop app and run multiple sessions and agents at the same time as well. So much like Claude, you can also add skills to your open code.

24:15So what are skills? Skills are reusable sets of instructions. So if you're running certain commands over and over again, it might be useful to turn it into a skill.

24:26So let's try to add that skill inside OpenCode. Can you help me create a skill for Xcode refactor patterns and help me add it to the skills.md?

24:37And now it's created a refactoring skill which is gonna help clean up my code and make it run better. To run the skill, would need to restart. I'm gonna close this, come back to open code.

24:49Yeah. Let's try it. Can you run the Xcode refactor pattern skill?

24:53Alright. Turns out I'll I added an extra s to the MD file. It needs to be skill and not skills.

24:59Don't make the same mistake I made. So now it's recognized my Xcode refactor pattern skill and is now running the refactoring.

25:10Pick one for me and run it. And there you have it. It's running the skill.

25:14I wanna show you our commands. So what are commands? When you tap on the slash button, all of these options are commands, and you can add to them.

25:24So I think a useful command for me is git commit and push. So I'm going to try and add that. So I'm gonna copy open code dot a I doc slash commands.

25:35Using this as a reference, can you add a command called that git commits and pushes for me?

25:44Let's see if we can add the command. So we successfully added the command so let's give it a shot. I open up OpenCode.

25:50This looks for git git push and there you have it. I have a new command.

25:57Now I'm gonna show you an advanced feature. Imagine you can run OpenCode on your mobile phone just like this. I'm gonna show you how to do it using Termius.

26:06Let's get started. So to run open code on your iPhone, you need to set up your Mac as a server where it's running the open code. Then you need to download Termius, the app, and connect VISSH to basic run open code on your computer but displayed on your iPhone screen.

26:23Now in this demo, I'm just gonna show you how to make it run when you're on the same network within this in the same house or in the same office.

26:32If you wanna run it remotely while you're out and taking a bus ride or driving a car, you'll need to set up more stuff to create a split tunnel so then you can connect remotely to your computer. But that's going to be another video because it's not easy at all.

26:49And if you're interested, stay tuned for my next video where I'll go through how to use Cloudflare to do that. But to get started, you need to set up SSH connection on your MacBook. So first thing you need to do is go to system settings, and then go to sharing, and you need to enable remote login.

27:07And then you need to add a user account. And when you log in, you just type in a name and the password you use to log in to your Mac. And now you've enabled remote connection to your Mac using SSH.

27:18You need to download the Termius app. So on your iPhone, search for Termius and download the free app.

27:25After you've downloaded Termius app, open it and go to hosts. Click on discover local devices. You should see your Mac here as long as you're connected on the same Wi Fi network.

27:37Click on it. Click tick. Enter the username and then the same password you used to log in to your Mac.

27:43You should see a terminal, and then all I need to do is type open code and boom. We've got open code running. Like I said, this only works if you're on the same WiFi network.

27:52I will produce another video to show you how you can connect remotely when you're outside traveling on the bus or whatever using Cloudflare and split tunnels.

28:01Stay tuned for my next video. I hope you enjoyed this video and gained a better understanding of OpenCode. If you enjoyed the video, please like and subscribe to my channel.

28:10If you wanna learn more about AI, feel free to join my free AI community. You can join my community for free inside the description. I also have a vibe coding course.

28:19If you wanna learn how to code, you can also find that inside my community. See you inside.

The Hook

The bait, then the rug-pull.

Keith opens with the line every over-tabbed developer already knows is true: one terminal, one task, one spinning cursor while three other things wait. In twelve seconds he names the real bottleneck — attention, not compute — and introduces OpenCode as the fix. What follows is the most thorough walkthrough of the tool currently on YouTube: install to mobile SSH in under half an hour.

Frameworks