Cursor 2.0 - Full Tutorial for Beginners

In this video, I'll give you a full beginner's guide to using cursor 2.0.

This is an Aentic AI code editor. It is

a fork of a popular editor called VS Code, which means it shares all of the same core features except it adds AI on top, meaning you can generate massive

amounts of code, files, projects, etc. using AI models. You can use this for free. You can also pay for premium

free. You can also pay for premium subscriptions or you have more credits and you can just generate more code. So,

with that said, let's get into it. Okay,

so first of all, I'm going to assume that you've downloaded Cursor on your computer. If you haven't, go to the

computer. If you haven't, go to the Cursor website, simply download it, and then sign into your account. Now, once

you open up Cursor, you may or may not see a view that looks similar to this.

You may see one that looks something like this. You may see something with

like this. You may see something with the pane closed. Don't worry too much about it. What we want to start by

about it. What we want to start by doing, though, is just opening up some kind of folder where we can start writing some code and we can store all of the code for our project. So

regardless what view you're seeing or what your cursor looks like because it will probably look different than mine.

Find this file menu here. Go to open folder and then simply create a new folder on your desktop and open it. So

in my case I had a folder called demo.

So let me just find it right here. And

actually I think I called it demo 2. So

I just pressed on it and did select folder. And then when I did that I'm now

folder. And then when I did that I'm now inside of this folder which you can kind of see from the top here. So all of the code that I generate will be stored here and I'll be able to find it later on.

Just make sure you don't lose all of your code, of course. Okay, now we'll go through the rest of the features here inside of Cursor in just a minute, but I do want to quickly note some of the new things that they released in case you've used this before because it has changed

a lot. So, Cursor has been around now, I

a lot. So, Cursor has been around now, I think, for maybe 2 years. It's very,

very popular and they've constantly changed the user interface, you know, added a bunch of features, made it better, etc. Now, about a month ago from when I'm filming this video, they released their own model called

Composer. Now the reason this is

Composer. Now the reason this is important is you can change the different models that you use inside of cursor. And previously you could use

cursor. And previously you could use something like GPT, you could use Gemini, you could use Claude, but now they have their own model which claims to be four times faster than kind of

those similar, you know, comparison models. And if we keep going down here,

models. And if we keep going down here, they also have released a feature that allows you to run multiple agents at the same time, which is very cool and it increases kind of the efficiency or the amount of code you can generate with

this tool. Overall, everything else, if

this tool. Overall, everything else, if you've used this before, is mostly the same, but they have drastically redesigned the interface, which is why I'm making this video because it can be very confusing to understand how to

navigate it. Okay, so those are the

navigate it. Okay, so those are the major changes. Now, let's go into cursor

major changes. Now, let's go into cursor and let's start writing some code. So,

first things first, you're going to notice that in the editor, yours might look different than mine. The reason why mine looks like this is because I've applied a theme. So if you want to change the theme of your code editor,

you can use the keyboard shortcut controlshiftp or commandshiftp.

Now this is going to open up something called the command pallet. Okay. Now

this is going to allow you to kind of get a shortcut to various different commands. And if you're looking for

commands. And if you're looking for something and you don't know where to find it, you can search for it inside of here. So if you want to change the

here. So if you want to change the theme, for example, you can type in theme and then you'll be able to go preferences, color theme, and then change it to whatever you want. So the

default one I believe is just like cursor light or cursor dark or something. I'm using monokai because

something. I'm using monokai because that's my favorite. But overall nothing changes. It just, you know, looks a

changes. It just, you know, looks a little bit different, right? I'll also

note that there's a bunch of settings you can play with that will also help you understand the tool a little bit better. If you press on this settings

better. If you press on this settings wheel right here in the top right hand corner, this will open for you the cursor settings. And keep in mind all of

cursor settings. And keep in mind all of these windows that you see on the screen, you can usually resize them all.

So if you want to make something a little bit larger, you can do that. And

if you see this tab here, it works just like a Chrome tab where you can drag it around. You can move it to different

around. You can move it to different places in the editor and you can rightclick it and pin it or split it to the left or open in a new window. So

feel free to, you know, make this look however you like. Everyone has different preferences. So there's a bunch of

preferences. So there's a bunch of settings here. We'll go through some of

settings here. We'll go through some of them later on. Just wanted to note that those do exist and they're worth kind of reading through. Okay. So once you're in

reading through. Okay. So once you're in this kind of view here, you're going to notice that there's a lot of different ways that you can view this editor. Now,

what they've done is they've centered a lot of the focus around this kind of agentic chat window, which is this right here, where you can type a prompt just like you would to something like chat GBT, and it can start generating code

for you immediately. So, we could do that, but I want to show you all of the different views that you have so you know kind of how to get between them, especially if yours looks a little bit different. So, you'll see that we have

different. So, you'll see that we have agents up here, and then we have editor.

Now, if you switch between these views here, you'll see that in the editor view, it will actually show you any files that you have. So, you can directly view the code like you would in a more traditional editor. Whereas, if

you go to the agent view, you'll be able to view all of the running agents that you have and kind of work directly with the model. So, this is a major change

the model. So, this is a major change that they've made, but you can just switch between them really easily like this. Now, you're also going to notice

this. Now, you're also going to notice that there's some of these kind of toggle buttons at the top left. So, for

example, if I toggle this left bar here, it toggles the agent sidebar where I can see all of my agents. If I toggle again, you see that now I'll get kind of like a code editor view and I can untoggle that

like that. If I go with this, I can

like that. If I go with this, I can toggle, for example, the terminal to appear here. If I toggle on the right

appear here. If I toggle on the right side, I can now open up my editors again. So, feel free to play around with

again. So, feel free to play around with these and kind of toggle or untoggle windows based on what you want. If we go back to the editor view, I know this is a lot here. You'll see that we also have this chat window here which allows you

to interface with one agent. But if you wanted to remove that and just fully view the files, you can press this button right here and it will kind of get rid of the AI pane. Okay, so that's the overall like UI outline. Don't be

too intimidated by it. Where we're

usually going to want to start is right here in this agent pane. So this is where we can actually start getting the model to generate some code for us. But

I do want to warn you that before you just start going crazy and asking it to do stuff, you really should have a general sense of what it is that you actually want to do. Do you want to build a website? Do you want to build a game? Do you want to make a back-end

game? Do you want to make a back-end tool? What is it that you want? What is

tool? What is it that you want? What is

the final outcome that you're looking for? The more that you can plan that out

for? The more that you can plan that out on your own and really understand what it is that you want, the better results you're going to get from these agents.

Because ultimately, these agents just try to guess what you mean based on the prompts that you give them. So the more information, the more context and the more discreet and specific you are, the better the results are. So first of all,

ask yourself what like medium do I want my application to be in? Is it even an app? Is it a website? Is it some backend

app? Is it a website? Is it some backend tool? Is it running on a server?

tool? Is it running on a server?

Understand that first. Then if it's something visual, you probably want to have a general sense of what you want it to look like in terms of maybe a wireframe or a mockup. And if it has some features, you probably want to know what those features are because if you

just let this thing go crazy, it will go crazy and it will generate a lot of stuff that is very hard to review and that you probably do not want. So

anyways, let's get into all of that after just a quick word from our sponsor. Now, you're going to see in

sponsor. Now, you're going to see in this video that AI agents can talk all day, but actually getting them to do something real is a whole different story and sometimes not even possible in

tools like this. Now, that's where Arcade MCP comes in. Now, it's the missing layer that gives your AI agents real power. Arcade MCP is a tool calling

real power. Arcade MCP is a tool calling platform that lets your AI securely act on your behalf. It can call APIs, access data, trigger workflows, all through a

unified SDK. Now, it takes you from an

unified SDK. Now, it takes you from an idea for a tool to running an MCP server to actually chatting with it right inside of Cursor or another MCP client in just minutes. So instead of manually

writing up dozens of APIs or worrying about permissions, Arcade handles all of that for you. Authentication, tool

registration, and security. So your

agents can just focus on getting work done. And the best part is it's open,

done. And the best part is it's open, simple to set up, and built for developers who actually ship real AI features, not just demos. So if you're building AI agents, co-pilots, or

autonomous dev tools, Arcade MCP is how you make them actually do things. Check

it out from the link in the description below and start building your first MCP server today. You'll be surprised how

server today. You'll be surprised how fast you can go from concept to working integration. Okay, so let's continue

integration. Okay, so let's continue here. And like I was saying, we want to

here. And like I was saying, we want to start generating a plan. So from this window here, you're going to see that you have agent plan and ask modes. Okay,

so what we're going to do is go to the plan mode. And I just want to quickly

plan mode. And I just want to quickly ask this to generate a plan for some kind of in this case, let's go website that we want to build. Now, I actually want to build a game because I think games are kind of fun. So, I'm going to

ask it to build like Tetris for me. Hey,

can you build a Tetris application for me and generate a plan for a simple front-end website that anyone could run in their browser that allows them to play the game of Tetris? Keep it really simple. We don't need any advanced

simple. We don't need any advanced features. I just want basic Tetris with

features. I just want basic Tetris with the score, the ability for the shapes to move around, and just like the most MVP version you can think of. So, what I just did there is I used a voice tool that I have on my computer called

Whisper Flow to dictate myself and just talk directly into the model. This also

has a built-in voice mode. It's not

available right now, but if I didn't have something, you could use the voice mode and kind of just input directly in the window. And I'm kind of streaming my

the window. And I'm kind of streaming my thought process and telling it what I want to do. And then I'm going to get it to start generating this plan. So, I'm

going to go ahead and press enter. I've

just left the default model as composer, which you can change down here to anything that you want. And you're going to see that it will take a second now and it will start writing up this plan for me. So you can see on the right hand

for me. So you can see on the right hand side it shows me the plan that it's writing. I'm just going to close the

writing. I'm just going to close the terminal here and it's giving me what's called a markdown file. Now for all of the plans inside of here typically it uses this MD file or markdown file. And

you can see that it's just generated this plan really quickly and it's came up with nine to-do list items that we need to complete in order. So now what we're able to do that we have the plan is we can view it by just pressing this

button, right? It's going to show us the

button, right? It's going to show us the plan. If we didn't like the plan, we

plan. If we didn't like the plan, we could ask the model to adjust the plan or change something around in there. We

also could manually change it oursel.

You can see that I can just type, you know, directly inside of this window if I wanted to do that. So, I would go here. I would review the plan, make sure

here. I would review the plan, make sure that it's good, and then I'm going to press this build button either from here or from here. Now, if you click this little down arrow, you'll see that you can actually build this in the cloud.

I'm not going to do that right now because that's a little bit more advanced. But for now, we're just going

advanced. But for now, we're just going to go build locally or just the default build. And when we do that, it's going

build. And when we do that, it's going to spin up a new AI agent now that's going to start actually executing these steps and building the plan for us, right? We're building this application.

right? We're building this application.

Now, while this runs, what we'll do is we'll just open a new agent so I can show you kind of the multi- aent feature and some other stuff that's interesting here. So, from this side, if you can't

here. So, from this side, if you can't see this, you can just toggle it, right?

So, toggle the agent sidebar. You'll be

able to make a new agent. If you go to new agent, same thing here. You're able

to write in your prompt and ask it to start doing something. You also see that voice mode button I was talking about right here. Now, what's interesting is

right here. Now, what's interesting is that you can change the modes, right?

So, we have a plan mode, which is what we used previously. We have an agent mode, which means it can just go and start building stuff and generating a bunch of code, or we have an ask mode.

Now, an ask mode is something that won't actually touch any files or modify anything. It will just give you a

anything. It will just give you a response, like a textbased response. So,

if you want to know how to run your project, for example, you can just ask it in ask mode and it will tell you how to do that. If you want to ask a question about something unrelated or you want to ask a question about a file, but you don't want it to change

something, use this ask mode is kind of the most safe mode that you have. If you

do use this agent mode, just be prepared that the model can make a lot of wild changes. And especially if you're trying

changes. And especially if you're trying to do small targeted changes, it might not always be what you want. And it can be a little bit messy to kind of fix that and revert it and go back to the things that has changed and potentially broke. Now, also, like I said, we have

broke. Now, also, like I said, we have the models. I'm just going to leave it

the models. I'm just going to leave it on the composer model, but you could change it if you want, but generally now just use the composer model. And then

you have a few other features like the ability to upload image, for example.

So, if you had like a mockup of what you wanted the app to look like, you could just put it here and tell it build this, right? You have this web browser feature

right? You have this web browser feature now, which is something new, which allows um what is it? cursor to actually use a web browser to test the application for you, which is really cool, and it will just automatically do

that as long as you have that enabled.

You can also add other tools, which we won't talk about right now. And then you have this context, which allows you to actually refer to specific files, folders, terminals, browsers, all kinds

of stuff. So, you can explain to the

of stuff. So, you can explain to the model more in depth what it is that you're looking for. So, I'll show you that in one second, but let's go back to this agent here. And we can see that I believe it has finished. and we'll know

that it's finished because we now have this awaiting review tab which tells us that we need to review the code that it's generated and see if this code is good. So there's a little review button

good. So there's a little review button here. So I'm going to press on review.

here. So I'm going to press on review.

When we do that, it's going to open up something called a diff. Now a diff is essentially a list of all of the changes that have been made by this agent.

Cursors kind of change the way that it's shown the diffs now. And you can see that it's in this kind of file. And we

can collapse them or open up the individual file by pressing into it. So

here I'm just looking at one file. I can

click between the files or I can go back to this view and I can kind of scroll through them and just see everything that's changed. So what I typically

that's changed. So what I typically suggest that you do rather than just pressing, you know, keep all is test your code first, make sure it works, and then press accept all. The reason for

this is that the way the cursor makes changes is when you see them here, they're effectively already made. So

they already exist in your files. the

code is already here and it's going to stay there unless you press this undo all button. This is kind of an important

all button. This is kind of an important distinction because a lot of people get confused by this. Once you ask the model to do something, the code it's generated is just there. Like if I close the file right now or close the project, reopen

it, this code is still here. If I wanted to remove it because maybe it made a mistake or it did something I don't want, I need to press this undo all button. but it can sometimes go away or

button. but it can sometimes go away or you can lose this undo all button especially if you get further into the project or you keep using this agent. So

that's why I'm saying before we press this keep all button we want to make sure that we actually review the code.

So in my opinion reviewing the code from here is a little bit difficult. I don't

really like this view myself. So what

I'm going to do is switch over to the editor view. Now, from the editor view,

editor view. Now, from the editor view, you're going to see now that in my opinion, it's a little bit easier to read the code, and you can actually view the different files that have been written in the left hand side here. So,

it's going to show you all of the files and the directory structure. So, from

here, you can review the code, press keep, write, or undo if you know what the code is doing. But what I'm going to suggest is that you actually run the project, see what the output is, and then ask the model either to make some

changes or to keep it if it's working.

So, what I'm going to do is I'm going to change the agent here to ask mode. And

I'm going to say, how can I test and run this code?

And I'm going to see what it tells me.

And notice that because I'm in ask mode, it's not going to generate any files for me. So, it simply says, okay, install

me. So, it simply says, okay, install live server and then rightclick index.html.

index.html.

Okay, so if I didn't know how to do that, I would ask it how to install liveerver. But I know that if you go to

liveerver. But I know that if you go to this extensions tab and you type live server. So this is extensions here inside of VS Code or inside of cursor you can install this thing called live server which I already

have installed. Once it's installed then

have installed. Once it's installed then I can go back to my files. Notice

there's some buttons up here right that I'm pressing that allow you to toggle between different features. See there's

a bunch of them. Extensions, source

control, searching, right? Probably

pretty intuitive. So I'm going to go back to the explorer and I'm going to rightclick index.html html and press open with live server just like it told me, right? So, I'm going to open that

me, right? So, I'm going to open that with the live server. And now this goes into my browser. I'm going to press start game. And I can see that I can

start game. And I can see that I can kind of move this around and that I can play Tetris.

Okay. And it seems to be working pretty well. If I press the enter key, I can

well. If I press the enter key, I can pause it. Okay. Pretty cool. So, that's

pause it. Okay. Pretty cool. So, that's

all good. So, because that's good, what I'm going to do is I'm going to go back here and I'm going to go to my review page. So, either I can do it right here

page. So, either I can do it right here by just pressing keep all or I can go back to agents. And let's just close these. And I can press keep individual

these. And I can press keep individual files or keep all. And I'm going to press keep all. And now I've kept all of the files. And now the changes that you

the files. And now the changes that you see here, it's just showing me the changes that were made, but I can't really like undo them very easily. So

that's why you want to be really careful when you do this review and you keep it or you undo it. I hope that kind of makes sense. But that is how you review

makes sense. But that is how you review the code, which is probably the most confusing part to be honest of this tool. So what I'm going to do now is

tool. So what I'm going to do now is show you a few other features that are definitely worth noting. So once I've made kind of some major changes, I do like to go into the editor mode. Now

from the editor mode, maybe I want to like read through the code and maybe I want to make a change to a specific function. You know, if I actually know

function. You know, if I actually know what I'm doing, right? I know how to code. So if I like highlight this

code. So if I like highlight this function for example, you'll see that I actually have a few uh little things that pop up directly for me. Now I have this quick edit which I can access with control K or command K. This will allow

me to just make a really small change directly in the file. So I can say you know clean up this code or something and add comments. Okay. And if I do that

add comments. Okay. And if I do that then it will just change whatever I've highlighted. So I'm doing like a really

highlighted. So I'm doing like a really targeted precise change. And then of course if I don't want that I can just reject it and get rid of it. Right. I

also have the ability to reference this directly in my agent. So maybe I'm going to hit this button, right? Control L.

And now it's going to reference this code directly. And I could ask the

code directly. And I could ask the agent, you know, what does this code do?

Okay, I just spelled that wrong. But now

I'm in ask mode. So it's going to read through this and it's going to explain it to me kind of line by line so I can understand what it's doing. And you can see it's quite fast. And if I didn't want to do that and maybe I wanted to just reference an entire file or

something, I can use this at symbol and then I can reference individual files. I

can reference documentation. So I could add like my own documentation here, for example, and reference that or I could just reference a file. So let's

reference like game.js. Is this, you know, good code? I don't know, random, right? I'm just showing you that you can

right? I'm just showing you that you can kind of add the context to the chat window. And these are useful things that

window. And these are useful things that I always do as a developer. I'm like

looking at a specific function where maybe I don't quite understand it or I don't know know why it was generated and I'll ask the model you know explain this or fix it or make this really targeted change. So at the beginning you start

change. So at the beginning you start doing a lot of changes and as the code gets larger and larger you do smaller amounts of changes that are a lot more targeted so you get a more predictable result that you're actually able to

reveal. Now, for some reason you are

reveal. Now, for some reason you are planning on actually writing your own code. I just want to quickly show you

code. I just want to quickly show you that if you do start writing, you will get autocomplete by default inside of cursor. So notice that it's just like

cursor. So notice that it's just like predicting the next line that I'm going to write. If you want to accept this

to write. If you want to accept this prediction, you can simply hit the tab key on your keyboard. By doing that, it will just complete the line. So you can see like I'm just typing something and

then if I hit tab, it just does the completion. Just wanted to show you

completion. Just wanted to show you that. That's another way to kind of use

that. That's another way to kind of use the AI. like we have the chat window, we

the AI. like we have the chat window, we have kind of that quick inline edit that you saw that was like you know the precise change to a highlighted portion and then we have the autocomplete if you are actually writing your own code

directly inside of the file. Okay, so

that's most of the usage of the AI and just keep in mind that I can run multiple of these agents at the same time if I wanted to and I can kind of have an agent running in the background while I go and do something else. So

let's actually do that. Let's open up a new agent. And just so you're aware

new agent. And just so you're aware here, you should make a new agent when you want to do a different task. So if I still wanted to continue working on my game and I wanted to just kind of go

based off the code that was just written here, I would do it in this agent, the one that currently exists, because it knows all of the decisions it just made.

It has access to all of this context. So

it's going to probably give me a better result if I put it directly inside of this agent, right? But if I want to do another discrete separate task, I make a new agent. And then in this agent, I ask

new agent. And then in this agent, I ask it what I want it to do. So in this case, I'm going to say the following.

Please create a simple landing page that is decoupled from the game that just kind of advertises Tetris and has a button that says play and then redirects me to the game where I can actually start playing. So I'm just asking it for

start playing. So I'm just asking it for a landing page. Again, this is a separate thing kind of detached from the game. So I'm putting in a new agent. I'm

game. So I'm putting in a new agent. I'm

going to go ahead and press on start.

And then it's going to start actually generating the code. Now, if I wanted to, I could have asked it to come up with a plan for this first, but because this is a relatively simple task, I've decided just to go ahead and press enter

and let it kind of do a more simple plan itself because it will automatically plan if it needs to and then start generating the code. Okay, so this agent's running. What I'm going to do

agent's running. What I'm going to do now is I'm going to go back into the editor and I'm just going to start doing some other stuff. So, for example, I'm going to open this. I'm going to close all these files. I'm going to start kind of cleaning up the tabs and I'm going to show you some other features that cursor

has. So if you go to the settings wheel

has. So if you go to the settings wheel here, what I want to show you is that you can change first of all all of the settings for like the agents for the tabs for the various models. You can

enable, disable, add your own models here. And you also have access to

here. And you also have access to various tools and MCP servers which you can put in here or you can go to rules, memories and commands. Now this is the one that I want to show you now. So

inside of cursor you have the ability to add project rules or global rules.

Global rules are also user rules. Now,

these rules are things that are going to be applied whenever you write code either globally if you're using a user rule or within this project if you're

using a project rule. Now, rules are really smart to set up if you want the agent to follow a consistent kind of format or pattern or you just have something you don't want it to do or that you want it to always do. For

example, if your boss tells you, hey, you can only write in this style of code or you have to use, you know, C++ 14, you can't use a newer version. You can

just make a rule in cursor and tell it, hey, always do this thing, always do this thing, follow this rule. And

without you having to repeat that multiple times, it will just automatically follow it. So, for

example, I can go to add rule for a project cuz I just want it for this specific project. And I'm going to say,

specific project. And I'm going to say, you know, let's go just general for the name of the rule file. And I'm able to just write in the rules here that I want to be followed. And you'll notice that there's this thing that says always

apply, apply intelligently, apply to specific files. So you can be really

specific files. So you can be really targeted with your rules. Like in a specific file, maybe you never want it to touch it. You can just do this and say, okay, you know, for any.tsx file, I don't want anything to be touched. I

don't want you to change anything. In my

case, I'm just going to go always apply.

And I'm just going to put a rule. And

I'm going to say always generate doc strings for functions. Okay. And now that I have

for functions. Okay. And now that I have this in the rule file, you'll see this folder gets created that stores the rule for me in this markdown C file, which is markdown cursor rules file. And if I

start writing code now, this rule will just automatically be applied. And it

should, if it writes a function, follow that rule. Now, a few other things to go

that rule. Now, a few other things to go over, but let's quickly go back and check our agent. So, I'm going to go to the agents tab, and we can see that this agent is awaiting review. So, it says that it's kind of done the landing page.

The landing page is here. I can press on review. I can see all the files. In this

review. I can see all the files. In this

case, I'm just going to kind of blindly press on keep all because I can just trust that it's done that. And then we can check out the landing page and see if that works. So, let's quickly check out the landing page. I'm going to go back to the editor again because this is kind of the view that I prefer. I want

to understand how to run the landing page. So, I'm going to say, okay, let's

page. So, I'm going to say, okay, let's go to ask mode. You know, how do I open the landing page? And then let's see if it gives me the instructions. And okay,

it says that I can just rightclick it and open it. So, let's go to I think it's index.html. Let's right click it.

it's index.html. Let's right click it.

I'm just going to open it with the live server. And you see now that we get the

server. And you see now that we get the Tetris landing page. And if I press play, it runs it and brings me into the game. Boom. Right there you go. All of

game. Boom. Right there you go. All of

that's working. Pretty simple for it to do, but you can see kind of how I'm following along with this process. Now,

let's keep going and I want to show you a few other features that are worth noting. So, first of all, version

noting. So, first of all, version control. Okay. So if you go to kind of

control. Okay. So if you go to kind of this view again, go into the editor, you'll see that you have this kind of like branch looking thing which is your version control window. Now if you also

go here, you can just press on source control or version control. Now this

uses something called git. If you don't have git installed on your computer, you will need to install it. And this is something that allows you to essentially track and store all of the code changes that you're making so that if something

goes wrong in the future, you can kind of revert back to a previous version.

It's kind of like checkpointing your code or like saving your work in a more permanent way so that if the model goes and like makes all these wild changes and it does something really really bad, you can at least get back to a state

that was working previously. Now, I'm

not going to explain Git in depth, but what I'm going to highly suggest you do is watch this video that I'm going to put on screen right here, which explains Git and how to use it. Now, even though

the AI can do the source control for you, I would highly, highly, highly recommend that you understand this version control concept because it will save you probably a lot of nightmares that will come up if you are building

especially a large project. Now, if you just want to kind of cheat through this, what I'm going to ask you to do is just pretty much tell the agent to use git.

So, I'm going to tell it, hey, use source control /git and save this work.

Okay? Okay, so that it's going to just run the commands automatically for me to essentially save the work for me. Okay,

and this is something I would suggest doing. So you can see that it's now

doing. So you can see that it's now showing me, okay, it's running this command. It's seeing that there's no,

command. It's seeing that there's no, you know, git repository here. It's

creating a git repository for me. And

this is something the model's capable of doing is automatically running commands for you. Now, sometimes it may ask you

for you. Now, sometimes it may ask you to approve a command that's intended.

You can approve it, you can reject it.

In my case, I have it set to just always allow. So it always runs the commands.

allow. So it always runs the commands.

And you can see that what it's done is it's initialized a new git repo, added the files, created a commit which is like a save or a checkpoint, and then actually just saved the work for me. And

now if I go to my source control, you see that this kind of new window appears. And in the bottom left, you see

appears. And in the bottom left, you see that it shows me actually the commit that just occurred. And the more commits that you have, you'll kind of get this timeline of all of the changes that have been made. You may not understand it

been made. You may not understand it 100%, but this is a really useful thing to use for your projects. and I highly suggest that you do it. So that's pretty much it. Just the last thing I will

much it. Just the last thing I will mention is that you can use MCP servers as I kind of briefly touched on before.

So you have something called tools/mcp.

You have the ability here to add MCP servers by pressing this new MCP server button. Now I'm not going to really walk

button. Now I'm not going to really walk through this, but you can see this is an MCP server that I have that allows me to run something called Tiger Data, which is a pretty cool kind of forkable AI database. And if you add different MCP

database. And if you add different MCP servers here, your agent now has the ability to call those servers and kind of connect to external tools and just has more functionality. So just kind of be aware that, you know, if you've heard

of MCP before, model context protocol, you can add that in here. In my case, I've got four tools in mine and you can see kind of what some of the tools are and some of the things that it can run.

You get the idea, okay? And you can kind of toggle it on and off. I'm just

mentioning it exists. Look into it if it's interesting to you. Not going to be in the scope of this tutorial. So that's

it guys for this video. If you enjoyed, make sure to leave a like, subscribe to the channel, and I will see you in another video.