Collaborating with Agents in Your Software Development Workflow - Jon Peck & Christopher Harrison, GitHub

[upbeat music] All right.

Well, let's, uh, let's get, uh, get started here. I did see a couple people, uh, walk in at, uh, kind of the last minute. Um, if you haven't already, uh, read the slide there, uh, you'll notice the slide has the, uh, first bit of instructions.

Uh, send an email. That email address that you see right up there, so [REDACTED:email_address]. And then just as the subject, just put in your GitHub handle. One quick note about the GitHub handle.

If you are using enterprise-managed users, uh, this is gonna need to be a personal, um, GitHub handle. So that email address, and then your GitHub handle as the, uh, as the subject.

Again, not... But you know what to do. So go ahead and, uh, and do that. And, uh, while everybody's, uh, getting settled in, uh, let's go ahead and get rocking and rolling.

So, uh, my name is Christopher Harrison. Uh, the [REDACTED:gender] over there is, uh, Jon Peck. And, uh, we are both, uh, enterprise advocates at, uh, GitHub. Uh, obviously doing a whole lot of, uh, developer relation stuff, so basically talking to customers and, uh, doing events like this, primarily focused in on developer experience in this little thing called

GitHub Copilot. So this is going to hopefully be a hands-on lab. Now, I am going to put out a couple of real quick caveats because we had the exact same lab this morning, and we ran into a couple of challenges, the biggest of which with the Wi-Fi network.

That because what we're going to be doing is going to rely quite heavily on the network, we're going to try, okay? I'm gonna repeat that one more time. We are going to try.

We are going to give it our best effort. If it works, wonderful. And if it doesn't work, here's what we're gonna do instead, is you will still have a day or two to be able to complete the lab, and you can go ahead and complete that on your own time.

And what I will do instead is I will start doing some demos. I will, um, answer a whole bunch of questions. Um, the morning group had a whole bunch of questions, and something tells me that this group will also have a whole bunch of questions.

Been doing enough Copilot talks, there always seems to be an awful lot of questions. More than happy to talk through all of that. So we are going to try our absolute best.

Hopefully, everything will work, and if it doesn't work, we've got a nice little fallback. And I wanna make sure again that I highlight you will have access to the lab, um, for a couple of days after this event, so that way you could go out and do it on, uh, on your own time from, uh, from there.

So one last time, because I did see a couple of people that just walked in, notice the instructions on the slide. Please send an email,

[REDACTED:email_address]. The subject, just put in your GitHub handle. Again, make sure that it's your personal or a non-enterprise managed user account. If you don't know what enterprise managed users are, don't worry about it 'cause then you probably don't have one.

If you have one, you'll know it. See a couple of knowing nods there. Okay. Without any further delay, let's go ahead and talk a little bit about GitHub Copilot.

So Jon's gonna be working in the background here. Once, um, um, he is done working, I'm then gonna go ahead and, uh, make an introduction and get everybody pointed at the, uh, at the lab.

But I wanna get started by talking about something that's very, very important when we're working with GitHub Copilot, and that is context. That you'll notice that we have a tagline for GitHub Copilot called your AI pair programmer.

Now, not only is it wonderful marketing, it kind of is, um, but it's also a fantastic description of how best to think about GitHub Copilot, that you wanna think about it as an AI pair programmer.

And the example that I always give, the analogy that I always give, is my partner and I woke up the other day, and she says to me, "Hey, I wanna go to brunch."

Fantastic. Who doesn't love brunch? And so I recommend that a spot, and she looks at me and says, "You know, here's the thing. Been there an awful lot lately.

I wanna go somewhere new." Okay, fantastic. And then I recommend that another spot. And then she says, "You know, now that I'm thinking about it, really kinda craving waffles."

Spectacular. Who doesn't love waffles? And then I recommend that another spot. Now, that's a very normal back and forth. That what happened was, is a request was made. I had some level of context, and I gave back a perfectly valid response.

And then some more context was provided, and then I gave yet another response, and that's a very common back and forth. And that may have been context that she had at the very beginning and maybe just hadn't yet shared, and it might also have been context that she sort of discovered as we were talking.

Again, all of this is very, very normal, and it's how we think about when we're talking about with other people, when we're talking with human pair programmers, and then also in turn with that AI pair programmer, that context is key.

Now, when we're talking about context, a lot of times we really focus in on the prompt when we're talking about context and AI. That I see an awful lot, especially with new developers to, uh, to Copilot, is they're really worried about, like, "Okay, well, how do I craft a prompt?"

And don't get me wrong, prompt crafting is an important skill. But context goes beyond prompt crafting. Because unlike a lot of other tools, which are maybe going to run all of your code through different parsers or maybe even build your code or potentially even, like, run your code, Copilot again behaves like an AI pair programmer.

It's going to be reading your code, just as if I brought somebody else new onto the team, and they're going to start reading your code. So making sure that that code is readable becomes very important.

So if you have good function names, good variable names, if your code is well-structured, Copilot's gonna be able to navigate that pretty well.

And if you're using a lot of single letter variables, if you're using a lot of abbreviations, Copilot's gonna come along and it's going to struggle with that, just again, as another pair programmer would.

In addition, comments. Now, we all know the impor- the importance of commenting code, and there is a strong part of me that has this feeling that good code doesn't need comments.

But here's the thing, is having a line or two of comments does help you make your code that much more readable.

And so even if it might be superfluous, if it's gonna make life easier, maybe just go ahead and add that in, 'cause it's not going to, say, slow down your application.

So having comments describing what it is that you're gonna be doing, how it is that you're gonna be doing it, is really going to help things out. And then finally, last but not least, is your project structure.

That having a good navigable project structure is also going to help Copilot. Now, the real takeaway that I want you to get out of all of this is this is sort of like just good code 101, that having good code is really going to help out an awful lot with context.

I'm gonna come back to, to questions a little bit later.

Okay. So how do we then help Copilot help us? And so number one, context, we talked about that. Number two, the intent. Make sure that you're clear about the intent.

What is it that you are trying to do? And be clear about that intent.

Now, one of the things that I see an awful lot of developers do, especially if they've maybe worked with like older chatbots, is when they start sending those initial prompts, that they're very terse, because they're used to dealing with older chatbots that maybe weren't really good at processing, uh, natural language.

And so it basically became like using command lines. That's not what this is built for. So be clear. If you want something to be done a particular way, and kind of getting into the specificity aspect here, tell Copilot.

Or as I like to say, don't be passive aggressive with Copilot. Now, sometimes I'll see a new developer, they'll go in, they'll, they'll, they'll give like a very short prompt, and then they'll go, "Well, wait a minute.

Why didn't Copilot figure this part out?" Well, did you know that part? Well, yeah. Did you tell Copilot? Well, no. Okay, well, tell Copilot. Like, if there's an important piece of information that you need Copilot to consider, go ahead and make sure that you tell Copilot that.

Okay. Now, Copilot supports a handful of different workloads. Um, you'll notice the one on the far end, Copilot Coding Agent, I'll get to that one in a minute. It's all bright and sparkly and so forth.

Um, that's not because that one's, um, necessarily best. Um, it's just because it's the newest, and as we all know, we need to just highlight in sparkles the, um, the, the, the newest item.

So, um, first up, let's talk about code completions. So code completion is sort of the original version, if you will, of GitHub Copilot. This is where I'm going in, I'm writing code, and then Copilot will suggest the next line, it will suggest the next function, and it will suggest the next class.

And this is wonderful. If you're in the zone, you know what you're doing, you're typing away, fantastic, and Copilot will be able to support you right there.

Next up is chat. And originally this was single shot, that I go in, I ask a question, and I get a response back. So maybe it can generate a single file for me.

But the problem with that is we think about even the simplest of updates. Like if I go to update a webpage, I'm gonna be updating an HTML, JavaScript, and a CSS file.

That's multiple files. And so this is where edits comes into play. And with edits, Copilot is now able to edit across multiple files. And then we take that to the next level with agent mode.

And with agent mode, now again still inside my IDE, what Copilot's gonna be able to do is navigate your project, figure out what needs to be done, and then start building code.

And it's able to work iteratively, and it's able to run external tasks, like for example, running your unit tests. And if it detects, hey, there's an error, it's even able to self-heal.

I'm gonna sneeze. [sneezes]

Bless you.

Thank you. Okay. I, I was waiting for the second one. I usually sneeze in twos. It'll, it'll, it'll show up in a minute here. Um, okay. Um, and then finally we get into Coding Agent.

And Coding Agent is different because this is going to be done inside of github.com. So I'll have an issue on github.com, and I can then assign that to Coding Agent, and Coding Agent will then work in the background,

figure out what needs to be done, be able to again run scripts, be able to self-heal, and then eventually create a, a pull request that you can then review and eventually merge into your code base.

And so this is what we're going to be exploring. I'm not gonna worry too much about these two slides here. There we go. This is what we're gonna be exploring in this lab.

Jon, how we doing?

We are doing splendid. I have two people left to add before, and then I will need about three minutes to add them all to Copilot.

Okay. But we can at least go in and start accepting invitations in the meantime.

You can go in and start accepting invitations. One quick note, um, Michael Pope, I don't believe I have your correct handle, so if you could send me a new one, that would be great.

Okay. And there was one person who walked in just a minute ago, so we'll need to get that sorted as well, but we can get there in a minute.

Okay. So, um, first part to the lab. Navigate to the URL that you see right there. So [REDACTED:url]- uh, H-O-L, hands-on-on lab. So O-C-T-O-H-O-L, Hotel Oscar Lima.

And then at the very top... And by the way, I'm assuming that you're logged into GitHub with the handle that you sent over. And then up towards the very top you should see Accept invitation, and then go ahead and select Accept invitation

Now, here's the most important part about today.

Silence means, hey, everything is fantastic. So if you're not seeing it, please let me know.

Not responding to this.

Um.

I'm pretty sure, yeah. We'll see.

Okay.

Oh, yeah. Okay, cool.

And then accept the invitation right at the very top. Okay. Yep.

Let's see.

Did you send the email?

No, I didn't get it.

Okay. All right. Um, I'll come back in a second. Um-

This box, ask Christy.

Hey, Jon?

Yeah.

Do they need to ask for their Copilot seat?

Um, they don't need to explicitly ask. I am just going through-

Okay

... the process right now of adding those in.

Okay. Perfect.

We didn't know if you forgot to click the button to ask, and we go back and do that again.

It doesn't matter.

Okay.

Yeah, he's, he's gonna be adding you in. Okay. Now, for anybody who did not send the email address, um, please go ahead and send an email to that address, and then with your GitHub handle.

I'm gonna leave that up for about another 15 seconds here, because I'm trying to keep things moving for everybody as quick as I best can here.

Um, the, uh... I can share the, uh, share the slide deck. Um, my handle is, uh, [REDACTED:username], and you can, um, basically find me on everywhere with, uh, with that, including GitHub.

[REDACTED:username]. [REDACTED:username]. G-E-E-K trainer. Um, uh, GitHub, at GitHub.

Okay. Now then, let me get everybody rolling here. So,

after you have accepted the invitation... Um, who has not yet accepted the invitation?

I don't know. The email didn't go through or something.

Yeah, same.

I feel like I've sent it. Um, it didn't send.

Okay. And then what happened when you went to [REDACTED:username]? And by the way-

Yeah

... I, I, I wanna make sure that I highlight something. You'll, you'll notice that more people raised hands. Silence means everything is fantastic. So if something is not working, please let me know.

Now, for people who, um, uh, everything is working fantastic, then here's what you can do, is go to that URL.

I'm actually going to modify this slide real quick. Is...

And if we could get phones to a silent or a vibrate, that would be wonderful. Okay. All right.

So, um, to get started with the lab, what you're gonna do, navigate to that URL that you see right there. So gh.io/agent-sdlc-lab. When you get to the part to create your repository,

the owner is going to be [REDACTED:username], and then the name is going to be your handle, again, replacing your...

See where that's going. And go ahead and start working on exercises zero and exercises one, and I will let you have at it. And then I'm gonna start going around and helping people get unstuck.

Yeah, same issue as him. I sent the email, but I don't have a accepted invite.

Um, uh, okay. Um, hey, Jon?

Yeah.

When, uh, when you have a minute?

Absolutely.

Okay.

I've got two people in the... Where you at?

Uh, right over here.

All right. Thank you. I'll be right there.

Okay.

I can't find the email. Sorry.

Did you send the email?

Yes.

Okay. Um, we'll go ahead and get Jon over to help you out in a minute.

Yes? Probably an email.

I've been thinking like this, and I have an email address.

Okay.

And, um, and I, I can't see any-

Did, um... Was it maybe there earlier? Do me a favor. Um, hit plus right at the top here.

Okay.

Um, hit new repository. Just... Let's just test something real quick.

Um, hit the, um, dropdown right here. Um, okay. We'll go ahead and get d- Jon over here in just a minute. Sorry. Yeah.

Yes.

Okay. All right. Now then, remember, silence means everything is hey, e- everything is fantastic. So if something is not working, please raise your hand, otherwise we are never going to know.

What's this repository name?

Oh, okay. So note the owner. So owner is gonna be [REDACTED:username]. Nope.

Oh.

So notice the slide.

Okay, sorry.

Yep.

I'll make-

And then the repository name is gonna be your handle. There you go. Yep. Hey. I mean, it's giving, like, giving it a second. Um, okay. Then hold on for just a minute.

Jon will be over. All right. Hi. It is blank. Is it okay? Oh. Oh, okay. So here's what we need to do. Hi. [laughs]

Um, yep. Um, nope. Um, so yep. Uh, perfect. So go ahead and hit Enter.

And so go ahead and scroll down. And you see the step zero here, set up the environment? Let's go ahead and just hit that And then now we're gonna follow the instructions from here.

Okay.

So you'll notice that you're gonna wanna open up the, um, that URL in a new tab, and then you're going to select, uh, use a template, and then create from there.

Thank you.

You're welcome.

Just say...

All right.

Just go directly to github.com/.

All right. How's everybody doing? I know there's a couple of people that John's going around to. Yeah.

Can we get a copy of the definition of the owner to be optimal?

Yes. Yep. Beautiful.

Uh, okay. Good. Then...

Okay. Uh, who else had a... Something added? All right.

Yeah, I sent the email. I just sent it.

Oh, okay. Great. Doesn't matter.

I'll do it. [laughs]

Living on the edge. [laughs]

I have.

Yeah. [inaudible] Nope. I, um... Here's what you're gonna do is, um... So, uh, go to... Okay. So go to, um, docs. [inaudible]

And then just scroll on down and, um, go to step zero.

And then now go ahead and follow right along with what this says here. [inaudible]

Yeah. [inaudible] Oh, okay. So, nope. So, um, we, uh, we need to go to the first part here. So navigate to that URL here. [inaudible]

So go to...

Both of you, Daddy?

Yep. So, um, real quick, as people are getting started, I want everybody to kind of notice the, the steps on the slide here. So step one is go here.

What you're gonna find there is a set of instructions for the hands-on lab, and then in step zero, it's going to have you then create your repository. When you get to that step, you're gonna set the owner to [REDACTED:username], and then the name to your, your handle.

So start here, start working through the lab, and then you're going to go in and, um, uh, and set that up.

Okay. So let's go into step zero. Yep. And then now you'll notice it's gonna have you, um, open that up in a new tab. It's gonna have you create from- [inaudible] Exactly.

Yep. And then you're all set. Don't sit down yet. [laughs]

Yeah, so start at zero. Okay. So here's what we need to do is, um, um... Let's go ahead and, um, let's just, um, go to, um...

Nope. Um, you see the, uh, the URL up at the very top there, the gh.io?

Yeah.

Okay. Um, nope. It looked like we, um, uh, didn't, uh, uh... It should be, um, SDLC. So let's go ahead and just try that again. So type in... Type...

Nope, nope. Stop. Yep, there we go. So gh.io,

um, slash agent dash SDLC. Nope. Nope. Here we have an SL. We want SDLC. Yep. Dash lab. Enter.

Software development life cycle.

There we go. And so now go ahead and scroll on down. And then you'll notice there, step zero. Let's go ahead and go right into step zero.

And then now it's gonna have you create your repository. So if we read through, we'll notice step one. We'll open that URL up in a brand-new tab, and then follow the instructions from there.

Okay. So, um-

Right here?

Yep. [inaudible] Nope.

Oh, create a new-

Create a new repository. [inaudible] Yep. Nope. Put it into... Right where it says up on the slide. Hit the dropdown, [REDACTED:username], and then the repository name will be your handle.

Leave it as private?

Uh, you can leave it as private. That's fine.

Okay. All right, welcome.

And then you'll see the...

Yeah.

I created a brand new repo.

Mm.

Not downloading it. Can I delete it?

Yeah, you can just go ahead and delete that one. And then... [inaudible] Okay, so all you have to do is go to settings.

Scroll all the way down. All the way. Yep, there we go. And then, um, hit that.

Okay.

And then you're gonna hit delete. And then you're gonna hit delete. [laughs] And you're going to, um... You can actually just copy and paste it.

Okay.

Um, and then, uh, yeah.

Okay.

Welcome.

Sure.

Yo.

Um, there's no explicit instruction for us to merge the DR after

Correct.

So leave it unmerged?

Yeah, we can leave that unmerged.

All right.

Yep, yeah, yeah. We're just going in and, and kind of exploring it. There's also, um...

Yeah, I'll get to that in a minute. Ask me that question again in, in five minutes. I don't wanna interrupt everybody, but it's, there's an important thing there, so ask me that question again in five minutes.

Yeah.

I opened that specific lab, but I don't see use template or anything. Maybe because I'm not signed in, or?

Oh, yeah. You need to be signed in. [laughs] There you go. [laughs]

Okay.

Yeah. I created the Codespaces. Perfect. The repo I've just created. Yep. This works, right? Correct. Yep. Yeah. And, and just kinda keep on following on with the, uh, with the lab from there.

What this is gonna have you do in a minute here is it's gonna have you open up, um, the, your repository again, and then you're gonna create a couple of issues and start playing around with coding agent.

Yep.

Um, there might be some Codespaces where the actual Git connection hasn't established yet, 'cause that happens lazy after the initial load.

Yep. Yeah. Well, and that's, uh, the, the lab actually says to send a, a hello, um, to like, um, um, kinda wake it up.

Oh, okay.

Yeah.

Great.

We, we, we had planned for that.

This means I can hit all of them.

I know. Keep rolling. [laughs] Yeah. I'm on step one. Fantastic. So, I mean, granted there's something I can't solve. I run out of premium requests. Oh. Hey, John?

Sure. I'll bring a laptop. There's two classes, right? There's phone problems. [laughs]

He's having the premium request error.

Ooh, fun, fun, fun. Okay. Um, so I made a pull request against your instructions repo, so see your typo.

Oh, thank you. [laughs] I type really, really fast. I don't type, I type accurately, but I type really, really fast.

There you go.

See, here's the thing about being a developer. You don't have to spell correctly. You just have to spell consistently.

Yeah, yeah.

Yeah.

And, uh, you can rely on the red squiggles.

Yeah. [laughs] Yes.

Yeah, actually, if you're, you know, if you're like in a, in a zone, and you already like finish out step one and wanna start getting into step two, you can go ahead and do that.

You'll also notice that all of the steps actually include like little optional exercises. So if you are, um, you know, uh, somebody who kind of like just moves really, really fast, um, there's other stuff to do in, in, in that lab.

So yeah. So go ahead and, um, kinda keep working through exercise one. Um, I'm gonna do, uh, an overview of exercise one in like just a couple of minutes here.

Yeah, sorry

about

that.

For Codespaces, no. So Codespaces is a separate product. The reason that we're using Codespaces is it's effectively a container in the cloud, and so it just eliminates setup. So like it's wonderful for just development in general, but especially something like this, having people install software, forget it.

So yeah. Christopher. Does that okay?

I also have the same question as John. How much does it cost to get these tri-party codes? Does it come out of the box?

So Copilot, uh, for Enterprise, which is gonna be like everything that we're using, is thirty-nine ninety-nine, um, a month, so forty dollars a month. Yeah. Um, uh, there's also a business tier which does not, um...

It, there's a few different features that it doesn't include, um, but one of the biggest is gonna be, um, access to, uh, to coding agent. Um, and that's at twenty dollars a month, or again, nineteen ninety-nine or whatever.

And that's for the enterprise?

Uh, yeah. Yep. Yeah. Yep. Yeah. Yeah. Yeah, I'm stuck on setting up your Codespace.

Um, hit refresh real quick. It's on? Yeah.

But I know it doesn't work. I don't know if the model works for sure. Um, do you still have... Oh, oh, there it goes. There you go.

All right. So, um, here's what I'm gonna do. I know that people are still working. Um, periodically, I'm just gonna come up here. I'm gonna kinda talk and highlight a little bit about a feature.

So this is basically, um, a real quick overview of what it is that you're doing in exercise, uh, one. You can have a seat. Make yourself comfortable. [laughs]

Pretend like you're an adult. [laughs] Do I look like an adult? No. [laughs]

Proof that you've worked with somebody for way too long. Um, [laughs]

no. So, um, so let's talk a little bit about, um, GitHub Copilot Coding Agent. So the whole goal of Coding Agent is to have a, um, an AI peer programmer, that what I can do is I can assign it a Copilot issue...

or, uh, sorry, a GitHub issue, and let Copilot begin work on it. Now, the first big thing about assigning an issue to Copilot, remember that whole speech we did about context, waffles, all that good stuff?

Yeah, yeah, yeah. So context. Context is key. And so if we take a look at the issue that we created, we went ahead and gave it a pretty decent amount of detail about what it was that we wanted to have accomplished.

So you'll notice that we're gonna be creating a, uh, functionality to allow for submission of games. So here's all of our requirements, and kind of list them all out.

And if this was gonna be something more robust, I might have even dropped a, um, a potential workflow that I wanted it to kind of step through as it was building out all of the project.

Again, don't be passive-aggressive with Copilot, that if there's some piece of information you want it to see and understand, absolutely 100% tell it about it. Now, the next thing is, after you assign it, it needs an environment in which to work.

And there is an, uh, there's a workflow that you can create, and it's called, he said, speaking slowly to stall, um, Copilot Setup Steps. And so when Copilot goes to do its work, it's going to be running inside of GitHub Actions, and it's going to, step zero,

run everything that it sees inside this workflow. And if you look through this workflow, what you're gonna notice, a whole bunch of steps to install different services and packages and so forth.

And that's now going to be the environment in which it's able to do its thing. Now, that also leads me to another very big point, which is about security.

Because here's what I'm gonna do, is I'm gonna take an AI agent,

I'm gonna assign it a task, and it's now able to go off and start doing things. And I wanna make sure that everything it's gonna be doing is going to be completely contained.

So the first big thing about working with Coding Agent is that it does not have access to the internet,

that by default, it is a firewall that's going to block it from any external resource. Now, you do have an ability to open up the firewall if you do want it to be able to access external resources, but by default it won't.

And that includes, by the way, installing frameworks and libraries.

So the only way it's going to have access to frameworks and libraries is either, A, you open up holes in the firewall, or B,

you set this up to install all the necessary libraries. Number two, this is running inside of GitHub Actions, and if you're not already familiar with GitHub Actions, one great thing about it is that it's going to launch an ephemeral space.

So it's going to set up a space, it's gonna do its work inside of there, and then it's going to collapse down from there.

So it will not have access to any other resources besides what's inside of there. Number three, it does not have write access to your repository, save for the branch that it creates.

So it cannot modify anything inside your code base. The only thing it's able to modify is just its branch. Number four, that branch, you're gonna notice, will not actually create a full pull request.

It's going to leave that in draft form, not become an actual pull request unless you come along and say, "Yes, I approve this." Number five, along those lines, if you are a good developer shop and you have different requirements, things like security checks, things like unit tests, end-to-end tests, regression tests, all of which need to pass, what

you're also gonna notice is that on this pull request, none of that can actually run...

Where's that button? Right there. Until you click that button that says Approve workflow runs. So again, trying to keep it completely self-contained. And then finally, last but not least, and this was a question that was asked earlier about reviewing it and then merging it in.

When it comes to the current version of Coding Agent, there is a restriction, and this is by design. Um, remember again, Coding Agent is brand new, so this can always change.

But this is by design, that if you assign an issue to Coding Agent, you are not allowed to then review its work. So you need somebody else to review its work.

So basically, it's sort of like in most organizations, you're not allowed to review your own work. You need to bring somebody else in. Sort of the same thing is that I assign the peer programmer, and then I need somebody else to come along and say yes, that this actually, uh, that this actually is okay.

All right. So questions on Copilot Coding Agent.

So let's say you are the only programmer, right? So for that exception, like this

That is a fantastic question. [laughs]

I want to review it.

And I'm just going to say that that is a fantastic question. [laughs]

Two IDs?

And, and, and I'm going to say that's a fantastic question. [laughs]

Um, other questions? [laughs]

Can you iterate on what it generates?

That's a great question. So the question is... That was also a great question. Um, so the question is, can you iterate? And the answer is yes. Is, um, what I can do, um, let me zoom out a little bit, that was a little obnoxious, um, is I can just go in and add a comment.

So if I go in and I add a comment and I say, "Hey, um-

What, what I potentially want it to do. Um, also add in an endpoint to search for games by name. There we go

And note here, these adding comments to pull requests does not seem to be working.

Thank you. And if I scroll back up, give it just a second here, and then we'll notice the cute little eyeballs again. So now, what's happening is that Copilot has started work again, and if I hit View Session, hopefully I've stalled just long enough here, you'll notice that it's actually started up a brand-new session.

So it's gonna take the work that it did, and then it's going to start back up again and begin work again from there. Yeah?

Does it work with self-hosted runners?

No. Not at the moment.

When?

I, I, I shouldn't have even said at the moment. [laughs] [laughs]

Let's, uh, let's, let's summarize that answer. No. [laughs] Uh, yeah?

Yeah. If, if you were-

I always feel like a press secretary whenever I'm doing this

... um, can you run Copilot, uh, on behalf of other organizations? Like if you have a GitHub app and you have other companies that integrate with it, can you run a G- a GitHub Copilot on behalf of them or do you have to just run that yourself?

So, um, give me a little bit more about what you're asking there.

Yeah. So I have a GitHub application.

Yep.

Um, clients install my GitHub application. It generates some rich prompts that we want to use with basically GitHub Copilot to run something on their code base. Can we do that?

No.

Okay.

So what you could do is you could stand up an MCP server, so that way they could use Copilot, talk to your code, and then start to interact that way.

But you're not gonna be able to have Copilot like do something to somebody else's code base.

Okay. So it's not like an open API that you can just have a Copilot stick in a code base and do stuff for-

No

... someone else. Okay.

Yeah.

Actually-

Yep

... the same question here. We have GitHub app and we're getting the pull requests in customers' accounts. So, so I think we cannot leverage like-

So yeah, so what you can do is if you have a pull request, you can assign that to, to Copilot at that point. But again, it needs to be on a repository that you own.

Okay. So if we, if they installed our app, so it's of course in their repo and we can technically assign the Copilot to do some work as well on behalf of the client.

Yeah. Right?

I don't know the answer to that. I'm, I'm gonna be honest, I, I don't know the answer to... 'Cause I don't know necessarily like what an app is able to do on, on behalf of the user at that point.

Um-

We should try it

... I, yeah, I would, I, I, I would need to get back to you on that. Yeah. My, my handle is [REDACTED:username]. Here, let me just,

um... Ah. Um, doo, doo, doo, doo, doo. There we go.

So just shoot me an email and I can try and dig up a, a more robust answer for you. Other questions? Yeah.

I have a question about integration. So what if like I have a self-hosted coding agent

Mm-hmm

... issue like, you know, I've already made the whole thing, and then how do I, like add additional things? Do I go and add as a comment or, uh...

Yeah, what if like the company basically does not what I want?

Yep.

Should I just give up on that and-

Yeah. This is a, this is a great question, and I'm actually gonna up-level this, um, a little bit. So when we're talking specifically about Coding Agent, if it doesn't do, um, something that, that, uh, that I wanted it to, or if it does get something completely wrong, and keep in mind that AI, just like human developers, is

going to periodically get things wrong, and it's sometimes going to get things spectacularly wrong, again, just like humans are going to sometimes get things spectacularly wrong. So, um, if that's the case, there's a few different approaches that I could take.

So one of them would be is just like I did, um, here a minute, uh, ago is, or three minutes, fifty-seven seconds ago, but who's counting? Um, is just add in a comment and just clarify, "Hey, this is really what I wanted to use."

So that's one option. Um, if it's like completely wrong, if it's just, um, completely out of left field, then the next option is you can just go back, um, all the way up to the very top here, um, and under Assignees, um, and I'm not going to remove this 'cause I, I wanna let it keep doing its

thing, but just unassign Copilot, change your issue, and then reassign Copilot.

And then, and then that way it will just completely start from, from scratch. Um, number three, and I think this is always, um, very important when, when we're working with AI, is if we see it get something wrong, one of the first questions that we really wanna be asking is why did it get it wrong?

And nine times out of ten, and I, I'm gonna come right back to this word yet again, is context. Is that it didn't fully understand not only just your request, but the environment in which it was going to be working.

So one of my absolute, um, favorite features, um, is, doo, doo,

um, Copilot Instructions. So on Copilot Instructions, what you're gonna notice, and this is read, um, in, um, in Coding Agent, and it's also read on every single, um, chat request...

Oh, how do I, um... Yes. I figured out a shortcut key on a Mac. Go me.

Thank you. [laughs] Um-

My GitHub AI is new.

Hey. [laughs] Um, but what you're gonna notice here is I have copilot-instructions.md. This is looked up by Coding Agent, and it's also looked at by every single chat request that I make inside the IDE.

And what you're gonna notice in here is I have a really nice high-level overview of what I'm doing and how it is that I'm doing it. So I've given it a little bit of introduction about the app itself and how it is that I'm building it, I've given it code standards, I've told it everything that's required before

commit, I gave it some global language guidance, and then I gave it some specifics on, on what it is that I'm building and how it is that I want it to be built.

And then if I scroll all the way down, you'll even notice that I highlighted that repository structure, so that way I'm clearly spelling out, this is where to go find things.

So this is a big place to go in and look if you want Copilot to have some additional context, some additional information. Um, another big thing that you can do, again, speaking to things that both Coding Agent and Copilot Chat are, are able to use, is to stand up an MCP server.

So this is actually one of the other exercises in the lab, um, is Model Context Protocol, and basically what an MCP server allows you to do... And I have a slide right here that shows you the, uh, the diagram.

I need everybody to memorize this diagram. There's gonna be a quiz on this later. [laughs]

Fantastic. Um, but basically, here's what happens. If somebody goes in, they make a request to Copilot, Copilot can then look through the list of commands that have been exposed out through different MCP servers, and then from there, go access external data.

Because this is always one of the challenges whenever I'm working with Copilot, or really anything that's built on top of an LLM, is that typically it's gonna be limited to just that training set, and that's it.

This allows it to access external, uh, information, and even perform operations on your behalf as well. So if you see Copilot Coding Agent doing things in a way that you weren't quite expecting, take a look at its context.

Take a look at what it has. So look at your Copilot instructions. Maybe put an MCP server. So if you've got, like, a really, um, uh, large repo or really large code base, and you want Copilot to be able to, like, better search and better reason over that, this is a place where, like, standing up an MCP

server, um, would come into, uh, come into play. Now, um, I do wanna highlight one last thing, um, uh, because Harold is in the room, and, um, uh, he would be very disappointed in me if I, uh, didn't highlight this.

Um, and this is .instructions files. Um, which... And I am not saying this just because Harold is in the room. This is one of the, of, of, of the things that, uh, that Harold works on.

Um, I'm not saying that just because he is here. This is true. This is one of my absolute favorite little features, um, where when we think about, um, what we're doing in coding, uh, is we're doing a lot of the same thing over and over again.

Like, I'm going in and I'm creating yet another React component, or in this case here, I'm creating yet another Flask endpoint. But I probably want all of those to follow a particular pattern.

And so what I can do is I can create a .instructions file, and I can then manually add this into chat, and this .instructions file can then have information about how it is that I want it to be created, the different rules around it, and then even point to prototype files that basically say, "Hey, you see that

games file over there? Use that. That's how I want this to be built. Hey, you see that test file? Use that. That's how I want it to be built."

And I can even have this happen automatically, that you'll notice that I have apply to header info. And so now, if it matches this pattern, so if it's inside the server test folder and starts with test underscore, it's gonna automatically apply that.

So let's say again, going to React components. I can then go ahead and set up a filter. So anything that ends with a JSX, uh, TSX, um, extension, I want it to apply this.

If you're a Svelte developer, like I am,

um... [laughs] All right. Okay. All right. I got a couple people. Fantastic. Beautiful. Um, then I can go ahead and set up a pattern so that anything that ends with .svelte is going to follow the instructions here.

Now, again, this part here, right now, only available inside of Copilot Chat, but again, other ways that I can start to provide context. That was an awful lot on context.

There was a question over here. Yeah.

How often do you update this as part of your software development life cycle?

Yeah, this is a great question, is, um, uh, how often do I, do you update this? Keep in mind that this is an artifact of your repository, and I very much want to treat that as such.

So, um, I sort of give this an analogy to, you know those little scripts that you spend like three days writing, but it's gonna save you five minutes a day?

Yeah.

That's basically what these are, is that you're gonna spend some time, and you really should. Create your instructions file. Create your .instructions file. Like, really make these robust, because what you're gonna wind up with then on the back end is increased productivity, but you're also then going to help ensure that the code that's being generated is generated

the way that you want it to be generated. So as things change, then do that. Yes, Mr. Harold?

Yeah. One of the... Is, anytime you add extra space, you wanna correct those immediately so that, so you get it on the, on the... When you start minimal and then you keep adding, and you, and you see it here in the option to, like, make it, "Don't do that," or, "If you're going to do this, do this

instead." That's-

Yeah. And I would even add, like, to that, that, you know, one of the things that's relatively unique about being in dev rel is I do an awful lot of demos.

Yeah.

And one of the things about doing a demo is you need that to work. And doing that [laughs] with, um, a probabilistic tool, like GitHub Copilot, um, does become a little bit of an adventure.

That will be the word that I will use there. Um, and instructions files have been probably the number one thing that I have leaned on to help ensure that I get some level of consistency.

So I really can't overstate the importance of instructions file. And again, I'm not just saying that 'cause Harold's in the room Yeah.

And Jon, I have an opinion here-

Yeah

... uh, which is also as important as who the [coughs] ... is, right? Like, yes, you have the person who's actually operating on the problem, but team leads need to

Yeah

... that part of your responsibility set. And as much as, like, your team makes a decision about what set of dependencies we're gonna use, your team's making decisions about what our patterns will be, et cetera, any of the architectural decisions.

Those should go into the instruction file.

Well, when you make those.

Yep, absolutely. Yeah.

So, uh, in your issue, you specified you want to pass the test, but it seems like that would be something that would be embedded within an instruction file to ensure that they're always passing.

Yep.

Is there any reason that it's here as a, a requirement in both, or does it matter? Or-

So, um, so the, the question is, is that, um, hey, inside your issue, you told it something that it probably should have been able to find elsewhere, that it was inside the instruction file, um, and, um, and you still typed it out.

Like, you know, do you really need to do that? Fantastic question. So I'm, I'm gonna answer that directly, but then I'm also gonna up-level that, um, uh, as well.

So to answer the question directly, I really didn't need to. Um, but this really, again, kind of falls under that category of, like, not being passive-aggressive with Copilot. Like, if there's something that I know, typing it out isn't going to hurt anything.

So adding that to that, if anything, it's just going to help it because now I'm just, like, further reinforcing the fact that, hey, this does need to, to, to actually work.

Um, now I am gonna up-level real quick because I think this also then brings up another question, which is, hey, now we've got the issue here. Um, and especially, like, if I'm working with, um, with Copilot Chat, where now I could have, like, a .instructions file.

And we've already seen that I can have those call other files, and then now I've got my Copilot instructions. And so that then kind of begs the next question, which is, well, is there a hierarchy here?

And the answer is no, that there isn't any form of a last write, that it's gonna consider all of that context and then make its best decision from there.

Which also then kinda leads to another thing, which we should really be focused in on, like Jon was highlighting with, like, our team leads, is let's try to be consistent.

So let's not start sending mixed messages because just like if we start sending mixed messages to our team and they're gonna get confused, the exact same thing is gonna happen with, uh, with Copilot.

Yeah.

Okay, just one other quick question. Sorry. On, on the MCP, uh, uh, diagram you had, you mentioned that you can't go outside the-

Uh, I can open up holes in a firewall.

So that specifically is a part of the onboarding of an MCP server, you have to open up the firewall, uh, as a part of that integration?

So, um, if I'm using MCP locally from inside VS Code, good to go. If, um, I'm getting Coding Agent to do that, then, um, when I add in the MCP server, I also need to open up the firewall.

All of it's actually really well documented, um, um, uh, to, to highlight all of that, but yes. The straight answer is yes.

Many of these MCP servers will give you op protocols, so adding that configuration becomes a little difficult. But it says a lot of places uses username, password basis is easy, but then otherwise it's really complex.

Yeah. So, um, so when you go in and, and configure that, you can, um, do that at the repo level. I believe you can do that at the org level as well.

And then what you do is you then set variables, um, and you can do that again at the repo level, or you can do that at the organization level.

So you can centralize all of that.

Can you review the code that it's, it's not... and does it start a whole new session versus, like, limited, like just review this file and make changes to this file?

Um, so I'm gonna... I, I, I'm going to inject my own question into that first, and then I'm gonna come back and answer that question. So the first question is, can you review the code?

And I, and I wanna make sure that I, I mention this because it's so incredibly important. Um, AI does not change the DevOps flow. That if I think about how I wrote code before AI, um, I would write code off my own memory.

I would be, um, writing code based on code that I'm seeing elsewhere. I would write code by copying and pasting from Stack Overflow, making a couple of changes, and then just crossing my fingers and hoping and praying that it works.

Or maybe that's just me. [laughs] Or maybe I'm just the only one willing to admit it. Um, so in, in that situation, what do I have? I have a DevOps flow where I'm gonna get Jon to come along and, and review it.

I'm going to have linters that are gonna come in and check to make sure that I'm following all my standards. I'm gonna have security scans that are gonna look for any potential vulnerabilities.

Then I'm... And I have all of these different things, all of these, these, these safeguards to help ensure that the code that gets into my code base is secure and is written the exact way that we want it to be written.

When I'm using AI, all of that still stands. So, um, I want my manual code reviews. I want my security checks. I want my linters. I want my unit tests.

I want all of that to go in and run. Now, to get specific to, um, the, the second half of your question there, which is, hey, it's just this one file that I want you to look at.

Um, when I'm working with Coding Agent, um, or really Copilot in general, there's always a non-zero chance that it's not going to follow what you want it to do exactly.

So if I say, "Hey, go in and update this file," it, it might work specifically on that file and really, in p- in particular, when I'm working about, with, with agent mode, it might work with that one file, but it also might do a couple of other things as well, and there's not necessarily a way to guarantee

that it's only going to do just that. If you want it to only ever do just, like, that one particular file, this is where, like, edit mode comes into play.

This is where ask mode comes into play inside my IDE because in those situations, I'm steering rather than letting, um, um, uh, than, uh, than, than letting Copilot steer.

Sorry. I just... So when you review the code, it's gonna start a whole session, and then you might get a different code with your comments not fully taken?

I, so there's always a chance that, um, when... A- again, Copilot is probabilistic, so there's always a chance it's not gonna give me the code exactly the way that I want it to.

But remember that at the end of the day, what this has done is this has created...

Just open this up inside of a brand-new tab. This has created a brand-new branch. And so I could go in and add in a comment and tell Copilot, "Hey, I want you to fix this."

But this is also a branch, so I could then just go in and say, "Hey, let me create a brand-new branch on that space," or I could go in and, um, uh, and then just clone that, and then just check out that branch, and then go in and manually make the changes as well.

And so, like in the situation that you're describing where, hey, maybe everything looks good, but it's just this one file that I need to modify. Honestly, in that situation, I think what I would do is I would just go ahead and, um, I would just do this, is, um,

get checkout. Yes. [inaudible]

mention that it's like needing to go and remember every last detail of every command.

And Copilot's great at generating get commands.

It is. [laughs] Um, so, uh, yeah. So I could just go in and do that. [laughs] Now I can just modify it manually. Or, like I could just work with Copilot here then on whatever it was that I, uh, that I was wanting to do.

Yeah.

Yeah. So can you go back to the diagram of the MCP?

Sure.

That has the big connect your computer section to that. So if you're using one of these remote agents and it's connecting to MCP, how... Can it do that, number one, with the remote session?

Yes.

How long does that session look like on your behalf?

So, um, so the... I, I'm gonna kind of up-level all of this. So the question is, is can it go to a remote server? Um, how long does that session last?

Um, et cetera. And the, the short answer to can it access the remote server? Yes. Can it act on your behalf? Yes. How long is that gonna last? That depends.

Um, that, um, when you go to set up that connection to that remote server-

Right

... you're going to provide a credential-

Yeah

... um, for that. So in the case of, say, like the GitHub, um, um, MCP server, then I can go in, and I can provide a token for that, but then I can set an expiration on that token.

Yeah, for sure.

Yeah. So it's... The... All that's really going to depend. There isn't anything in Copilot that's, um, going to, um, automatically, like sign you out or anything like that.

And so this MCP server, it can't basically kick off... You know, this is something that the actual, the agent that's running remotely has to decide to pull from. It can't basically kick off, um, the agent itself.

Correct. Yeah. So the, uh, it's, uh, the MCP server's not just automatically gonna be able to do something on my behalf. I'm going to have to indicate, "Hey, I want to do this."

And then on top of that, um, and I don't think I've got it set up, um, on here, and it would just take me a couple of minutes to do it.

Um, but if I go in and I make a request that's gonna kick off an MCP server, I actually have a button that says, "Hey, do you wanna do this?"

Which actually leads me right into my next point, which is, um, that my... I, I wanna make sure that I'm vetting my MCP servers just like I would anything else that's going to access external information, just like anything else that's gonna be able to act on my behalf.

Yeah. This is really important. Can, can I ask one follow-up?

Sure.

Is, is it possible to create a GitHub issue on behalf of another company, and then have that be assigned to their internal Copilot?

We're getting into a lot of specifics. Let's talk afterwards.

Okay.

All right. Yeah.

Um, do you have visibility as to what model the coding agent uses, or visibility as to how it's making its decisions so we can further profit in the future to help clients further?

So for coding agent, and again, I'm, I'm just to, to make sure that we're clear, this is where I'm on github.com. Right now, we're not exposing out the model that's being used internally.

Um, and there's a reason for that, um, because we have a very specific type of workload, a very specific set of scenarios that need to, um, to be accomplished here.

And so the model that we're choosing is the model that specifically... Like that, that we've determined is best for doing this. There might come a time where that starts to get exposed out, and maybe that, that allows for change.

But as of right now, it's, it's specifically because it works with this. Yes, you still have your hand raised.

One more quest- quick question.

Yeah.

Really good session. Is it possible to have more than one coding agent?

Um, yeah. I mean-

Like a pair of them. Like, can you make... Assign them to two people? Like, I guess it would cost more, but-

Yeah. Like I, I, I could create 100 issues and assign all 100 of them to, uh, to, to, to coding agent.

Oh.

Yeah.

So you could just do them... You would spawn up an environment-

Yeah. So remember that each time that you assign an issue to a coding agent, it's gonna create an environment specifically for that issue.

So each issue will get its own agent.

Exactly.

Oh, okay.

Yep. Yeah.

Those instructions, right, in that apply to section-

Yep

... seems like it's gonna grow so huge. Can I do it as, as opposite, as like in that folder have my own instructions and tests?

Override. Yeah.

Yeah, kind of like a chain and hierarchy. Yeah. Instead of like having one gigantic file.

Um, so, um, what you can do, and I actually have an example of that right here.

Multiple teams and then-

Yeah. So you'll notice that right here, um, I'm actually, um, inside this .instructions file, I'm actually referencing out to another .instructions file.

Folder. Okay.

Yep. And then on the apply to, you could go in and start setting folders as well. So however it is that you want that to be set up.

Okay. That's good. Thank you.

Yep. Yeah. Yeah Sure.

So, uh, is that possible for the 13 Copilot to access some, uh, you know, like HTTP sources with all these functions, like, I'm not... I'm, I'm working on big enterprise that where, where I need to get a pattern for both, uh, both patterns for everyone else?

Yeah, so if you're working with Coding Agent, you can configure your list of MCP servers at the organization level or on individual repos. For, um, um, for anything that's gonna be inside of, uh, VS Code, and I'm not sure how others would, um, would necessarily be s- um, have that set up, um, you can actually just set

up inside of a repository, or inside your repository, your mcp.json file, um, which is what I've got right here. And then I can just list them all here. That now becomes part of the, uh, the repository.

Um, the user will then still need to set up their own credentials because, again, it's going to act on, on that user's behalf.

But, but this guarantee not only for two, right? I can add to resources and-

Correct.

Okay.

Correct. Yep. Yeah. Mr. Harold?

The Coding Agent only supports tool right now. Uh, resources just landed into the scope.

Oh, okay. Thank you.

And, uh, for, for my talk tomorrow.

What time is your talk?

Oh, it's, uh, yeah, it's usually like 1:00 PM.

Okay. [laughs] [laughs] Yeah.

Do we have to worry about namespace collisions at all? Like MCP 207 says

So, you know, I'm actually going to, to kinda up-level this. Um, one of the things about using MCP servers all up with agent mode is that Copilot does need to figure out which MCP server to call, um, and when to call it.

And generally speaking, it's gonna be able to do that on its own. But again...

And, and, and let this maybe be the number one lesson that we pick up from, from today, that if I ever see Copilot not doing something the way that I expect it to do, what am I gonna do?

Write some instructions.

There we go. I'm gonna... Well, or, or even, even more generically, I, I... You know, back... I used to be a tech trainer. Um, um, so I would, like, literally just, like, go around different companies, teach classes.

And one of the things that I would do is I would have little, um, um, little bits of candy, um, and just, like, toss them out, so, like, the little baby Reese's, um, and toss those out, and I wish I had one.

Um, but, uh, but yeah, exactly. Like, update your instructions file or, again, like, improve your context. And so, like, one of the things that you could do is if you know for this specific type of scenario, this is the MCP server I want you to talk to, put it inside your instructions file.

And in fact, Azure, um, their MCP server, when you register it, VS Code will actually prompt you and go, "Hey, do you wanna add that into your, um..." [laughs]

Here, there, there you go. [laughs] Um, but, um, uh, but if you install the Azure MCP Server, it will actually say, "Hey, you should add this to your instructions file."

Okay. That was a lot of questions. [outro music]