Video: Vanta MCP Builder Workshop | Duration: 3488s | Summary: Vanta MCP Builder Workshop

Transcript for "Vanta MCP Builder Workshop": Hello, everyone, and welcome. I'm so excited you're all here. I might just give us one more minute, while we wait for folks to join. While we're waiting, I would love for you to introduce yourself in the chat. Tell us, where you're joining us from, in the world, and if you feel comfortable, what organization you're joining from. I would just love to see who's here. Awesome. Great. I see lots of people typing chat. Wonderful. All the way in The UK. Great. Alright. I'm going to give us one more minute, and then we will go ahead and get started. If you're just joining, we're just kinda giving everyone a chance to settle in. Feel free to introduce yourself in the chat. Tell us where you're joining from. We'd love to know who's here and where you're at. Some of you already recognize that Jaques is here. He is our program manager of down market education, leads most of our trainings, and he's here to support as well as Shannon, our senior manager of customer education here at Vanta. So we're we're happy that they're here. You might recognize them. Yes. Shannon posted a question. Just wondering if you all have a doc documents tab next to the chat modal. It should say, like, chat, q and a, maybe docs, files. Are you able to see that? Okay. No docs. Alright. We'll we'll figure that out. Alright. Let's go ahead and get started. I don't wanna waste any more time. Hello, everyone. If you're just joining, this is Build with Vanta, our MCP builder workshop. I'm Geralyn, the program manager of technical education here at Vanta. I'm thrilled to be here with you today, and we're gonna dig into our hands on session for connecting, exploring, and building with Banta's MCP. This session is special because it's the very first one. So this workshop will be in our regular training rotation starting in August most likely, But you all are really getting the sneak peek and VIP sessions, so we're really glad that you're here. Over time, this session will continue to change and evolve as our MCP capabilities do. So you can always check back. You can come back if you want a refresher, you wanna see what's new, how things are changing as our capacity changes. I already called this out, but some might have just joined. So there are some additional Vanta customer education folks here in the chat to help out in case you run into any questions. So, feel free to use the chat early and often. As the name implies, this is a hands on workshop style educational hour, so participation along the way is readily encouraged. I will just be demoing or presenting slides the whole time. Also, just kind of as an FYI, I have a few different screens that I am running here and looking at. There will be times I'm on camera, off camera, looking at these other screens or windows, but I'm here with you every step of the way, and you will not, move through the workshop alone. Alright. Before we jump in, just a note on our theme and what I want you to keep in mind. So this session is about AntaMCP, and, also, it is about finding a way to connect once across your tools and build everywhere you need in order to make your GRC program work better, more effectively, and more efficiently. So let's go ahead and jump into our agenda. While I go through this agenda, if you're just joining, would love for you to take a second to introduce yourself in the chat, share where you're joining from, if you feel comfortable which organization you're from. I'm joining from Baltimore, Maryland, so let me know if other DC, Maryland, or Virginia folks are here on the call. During our session, we're gonna do a quick MCP catch up to ensure we're all on the same page. We'll have a quick moment for connection and verification, then we're gonna dive into a guided exploration, Avanta's MCP. And from there, once we're feeling a little bit more comfortable, you'll do a build sprint that aligns with that comfort level. And finally, we'll hopefully have time to share a bit about what we tried and wrap up. Okay. Thanks. It's so great to meet all of you in the chat. I see the messages coming through. People from all over. That's wonderful. Alright. Let's go ahead and dig in. So before we do anything else, I want to show you something. Alright. Let's open up with a fun prompt here. If this takes too long to think, I do have one that's prepopulated, but we're just gonna give it a second, to to think here, and then we'll chat about it. For those that can't read the screen, if it's too small, my prompt says, find my top three highest priority filling controls in Vanta right now and post a summary of them to the CED team channel in Slack. Let's see what it does. We'll have to exercise some patience. I do have one I can jump to here in a second if it starts to take too long. Just so we know, the docs tab should be showing up for all of you now. So we will have some resources there for you shortly. And I'm going to go ahead and use my pre populated prompt. So here we go. I ran the same prompt this morning. Using the Vanta MCP, find my top three highest priority failing controls, post a summary. As you can see, it went ahead and queried my entire compliance program and found the three most important things to fix and gave me a Slack message to share and use to the team. It could push that Slack message, directly for me if that is the next action I wanna take. My claw just knows that I don't like for it to post messages for me, so it didn't do that yet. But all of this is happening from one prompt here at the top. There's no dashboards, no copy and paste, no context switching between five tabs. And what you just saw was sort of the read side of this analytical prompt. Right? The same connection can just as easily kick off the next step. That could be drafting a fix, opening a ticket, updating a record. I'm not gonna do that live right this second, but I do want you to keep it in mind because you'll get hands on with exactly that later in the build sprint. A key thing I want you to walk away with today, this isn't really a story about Vanta talking to AI. It's a story about your AI tool becoming the connective tissue across everything you use. Vanta is one really important node in that graph, of course, but the real power is what happens when Vanta is talking to your other tools too. So if you already have other MCP connectors set up at your company, Slack like I have here in mine, Jira, wherever your team lives in. You're gonna have a real opportunity later in the session to build something exactly like this for yourself. So let's hop to it and get going. This session is meant to welcome any and all MCP users. So we might have some folks in the room who are already using the Vanta MCP regularly, some of them might not be as comfortable yet. Let's all just start with the same baseline understanding. MCP stands for model context protocol. It's an open standard that lets AI tools like CLAW or Cursor talk directly to software. In our case, that software is your Vanda account. And as you just saw, it can talk to other tools too. Instead of logging into five different platforms, you just ask your AI tool. It queries and acts across your tools in plain language. Obviously, very simple in concept, but incredibly powerful in practice, which you just saw first hand. So let's get everyone connected and get this session started. And one thing to call out as we keep going. If there is ever a time where we are using a prompt that isn't working for you or isn't return anything, maybe simply because you don't have that area of the product in use or configured, you can use our help center article with MCP prompts and pick something that works better for your program. That article looks like this. So there are lots of prompts here. As you can you can see on the right hand navigation, you can click to a prompt area that works better for you. So feel free to plug and play as we go just depending on what's gonna work best for your program. Alright. If you had a chance and completed the prework and you're already connected to Vanta's MCP in your AI tool, I'd love for you to drop, you know, a thumbs up into the chat, maybe a green check mark or a, yes, I'm already connected. I'd love to see how many of you came prepared and already did your connection. K. Thanks, Charlotte. Great. Love to see it. Alright. Thank you. I know we have more than two. Okay. Some other people live reacting. Oh, no. Chris, it's okay. It's okay. We all have very busy work lives, and if it didn't happen, there is no consequence for not doing the free work. So if you did do it, great. If you did already, you know, feel free to use this moment to just, like, take a sip of your coffee, and we'll get everyone else set up very quickly. So if you didn't have a chance to connect, that is totally fine. We are going to, take a second and let you get connected. So the connection guide is in the docs section of our webinar tool here. So if you click on the docs tab and you go to connecting to VantaMCP, that link there will take you to this article you see on the screen right now. The connection guide walks you through the setup very specifically for Claude, Cursor, and Perplexity. However, it also explains, how to connect to other MCP compatible clients if you're using something else. So find this section for whatever tool you're using and follow those steps. I'm not gonna, like, model how to connect because, obviously, it really varies depending on which tool you're going to do. But, I'm gonna use Cloud throughout the session today. It only takes a few minutes to get connected. So there are some customer education folks here in the chat standing by, ready to help. I'll help as I can. But I'm just gonna go ahead and pause and give maybe, like, one to two minutes. If you're not connecting today, which I see a couple people in the chat saying, that is 100% okay. You can still watch, learn, engage. It's totally fine if you either don't wanna connect right this moment, maybe can't connect and are having issues, and wanna come back and do it later or might be waiting on a review in order to connect. But let me go ahead and pause here. And if you're able to get to connect, we just haven't had time yet. Let me give you, you know, one to two minutes to do that. Alright. About one more minute, and then we'll keep moving. If you're just joining right now, we are just having a couple quick minutes to connect the VantaMCP to our AI tool of choice using the link that is in the chat or in the docs tab, and we'll come back together shortly. Alright. Again, if you weren't able to connect right now for whatever reason, please don't worry. You can still follow along during the session and learn, engage, and then try it out again later when you have more time to connect. But for those that are connected, let's go ahead and run our maybe your first prompt ever or maybe just a prompt that we can all do fairly quickly. This is our hello world moment. Let's make sure everything is working well. So I'm gonna jump back over to Claude, and let's type in our prompt. We're gonna say list the integrations connected to my Vanta account. List the integrations connected to my Vanta account. Let's run it. Let's see what we have connected. Go ahead. You should be doing this, as well. Go ahead and post that prompt in whatever AI tool you're using, whether that's Claude, Cursor. And, I'd love to see in the chat if you can share maybe a couple of integrations that it's returning to you. Let me know if you're getting the response. This prompt happened a lot quicker than my opening one. Here, it has already listed out the integrations that I have connected, which is awesome. What are you all seeing? Are you getting your integrations returned to you in your response? Are you able to see what you have connected? Alright. I got one thumbs up. Awesome. Thank you. I'd love to hear from more. If you feel comfortable, you can drop your integrations into the chat or just let me know. Yes. That worked. And now from this, of course, you can spin this out into follow-up, prompts. Right? You might not see something on there that you expect on there, and you wanna reach out to your IT admin about the integration, and you wanna send them a Slack message. Maybe you wanna open a ticket. There's all sorts of follow-up actions to this. Right? But if you are seeing your integrations pop up, it's working. You're officially connected, and we should be go good to go to keep moving. Alright. Thanks for responding. Thanks for letting me know you are all in. So now that we're connected, let's see what this thing can really do. Right? Everything we've done so far I'm just checking the chat real quick. Also, it's recommendation. So I don't necessarily have a recommendation to use SONNET, and that's just what I'm using. So you can definitely play around with it and figure out what works best with, you know, your preferred type of responses. I've also gotten that response not in a table format. Weirdly, when I use Claude desk top, it doesn't return it in a table. When I use Claude browser, it does. So it might just take some playing around with, to figure out kind of, like, what's the right cocktail of AI formatting that gives you what you want. Alright. Here is how this next block of our workshop is going to function. So this is a guided exploration. We're gonna do something similar to our integration one, but now these are gonna be a little bit more complex. So we're gonna run three prompts together one at a time. I want you to run it yourself against your own account while I do the same on my testing account, and then we're gonna drop observations in the chat. So it really helps when we are, sharing and engaging. So please feel free to drop a chat, you know, what surprised you? What's useful? What would you wanna dig into further as we move through the next section? There's no wrong answers. So the more that you share, the more fun this gets. So let's go ahead and look at our first prompt. Nope. Not all of them. That's gonna be overwhelming. So our first prompt says, what are the top five things I should fix in my Vanta account right now based on failing controls, open issues, vulnerabilities, and overdue vendor reviews? Just a quick note, this prompt can take some time. So you can shorten it if you want and just kinda pick apart exactly the type of response you would want, or you can run the whole thing and just wait it out. But I just wanted to call out, it could take a second. So go ahead and take this prompt into your AI tool, and let's see what it returns. And I can drop this one into the chat so you can copy paste it. Alright. And I'm gonna come over here and go to my first prompt. Alright. I prepopulated just because, I wanted to make sure I could, not be waiting, but I want everyone to have a chance to run it. So go ahead, take this prompt to your tool, and, in a second, we'll regroup and see what it returned to you. Alright. I hope yours is running, efficiently. So mine gave a prioritized fix this now list directly pulled from your Vanta account. Also, just to flag again, I'm using this test demo account that many Vantas use, and so it has a lot of issues. So it gave me top five things to fix. Critical open issues with no remediation started. I have some overdue vendor security assessments. I have some vendor risk data missing. I have some overdue issue backlog. So I have quite a bit going on here for me that it wants to flag. But it's even prompting me that it could start the next version of this this conversation, right, which it's asking if it wants me to pull the remediation guidance for any of these five, so maybe exact steps to help fix it or draft your tickets for the team to help address this. So it's already kinda prompting me into that next chain, which is ultimately what our conversation is about here today. In the chat, I'd love to hear, you know, did this prompt work for you? What's surprising to you? What would be useful about this? What would you wanna dig into further? You know, what's helpful or effective about this prompt? I'd love to hear from some of you in the chat. Yeah. Yes. That is that is worth calling out that it could still be running. That's totally fine. It can take a minute. No worries. Anyone else get their answer yet and willing to share? What surprised them? What's useful? Oh, yes. Great. That is a very good use case. Thanks for sharing that. That's also a very next, like, a useful next step which relates to, some of the builder prompts we're gonna get to here shortly. Yes. It's very helpful for calling out, some tasks we need to look at. Yeah. Great. Thanks, Kevin, for sharing. And here, of course, with the MCP, you're able to do that tool of chaining, right, which is what we're talking about today, where you're still able to think you know, to get the things you need out of Vanta to get the data and information. You can also take right actions within Vanta, but you're also able to connect it to the rest of your tool ecosystem, which is really helpful. Okay. Let's go ahead and move to the next prompt for our guided exploration, and there are two options here. So the first option is universal. This should work for everybody who's connected. This one says, look at my policy library and tell me which active framework requirements don't have a corresponding approved policy in place. And then if you're interested, you could use the vendor risk option. So this one says, look at my vendor list in Vanta. Sort vendors by inherent risk level and tell me which ones I should prioritize for review. So let me drop these into the chat for you. That way you can copy paste. Again, I would just pick one of these, either the universal, you know, if that will apply or if you know that vendor the vendor area of the platform is something you have configured and you wanna run that one, you can go ahead and do that. But let me just give you a second to try those out. While you are copying pasting that to run your next prompt, I'm gonna jump over here to show you what happened with mine. So this is the policy library one. I'm having it tell me which active framework requirements don't have a corresponding approved policy in place. Yes. Thanks, Eric. Sorry. I know the copy paste was a little strange there. There are two prompts embedded in that paragraph. So here we have two frameworks that have zero corresponding policies, not even drafts. It's also calling out frameworks with policies drafted but not yet approved. And it tells me my best covered framework, which is really helpful and great to know. So it's returning this one really well, and there's a lot of useful information I can take from here on what to prioritize and what to work on. So let me know if you've decided to run the universal prompt. Let me know what that returns to you, if this is useful, what surprised you, what you would want to dig into further. Super curious to see what you're getting. Is everyone's AI tool still thinking? Oh, great. Love to see that, Chris. That's awesome. Really useful information there. Yes. Andy, I know the it can take a minute. It can take a minute. Hence, by prepopulating all of these chats, I didn't wanna make everyone wait. Okay. Eric, great. Love to see that. Awesome. Did anyone run the vendor risk prompt? Anyone have anything to share? This is what mine looks like. So it had, it gave me a vendor risk lands landscape here and prioritized for review, listed some of those here, and already gave me already reviewed. It even gave me a recommended order of operations. So some really useful stuff there even on the the vendor one. Okay. Great. Yes. Okay. Any other takeaways before we jump into our last guided exploration prompt? Thanks for the emojis, reactions, chats. It's really helpful. Good to know where everyone is at. I think this one is always really interesting because people often discover vendors they forgot about. It kind of can bring some stuff back up to the surface that you might not have noticed, just from, you know, the tool itself. Let's run our last prompt. This one is sort of like the building. Right? We sort of gone from, you know, we're we're moving down the line here on what we want it to return to us. This says, look at my failing and overdue test in Vanta and map them to the, insert framework here, controls say effect. For each at risk control, tell me what the control requires, which tests are failing against it, and how likely it is to result in an audit finding. Prioritize by audit risk. I'm gonna drop this in the chat since it's a little lengthy here. And, again, you'll have to fill in that, you know, if you wanted to return the SOC two or maybe HIPAA. Go ahead and edit that section there. And I will show you how this looked online. I'm just gonna give you a second to start the to start your your prompt though. Okay. So as your system is thinking, hopefully, let's look at what mine said here. So it's giving me critical risk. Right? So I'm having some two factor authentication issues here. Then I have some high risk things that I need to look at, and then I have some medium risk to consider and some lower risk. So it's also then at the bottom giving me that priority order to remediate, which I asked for. All of these two speak to part of the prompt where I was asking about how likely it is to result in an audit finding that's kind of how it's categorizing here. So this is really useful information. Right? It's really helpful for me to know what I need to prioritize, and it is, again, kind of looking at your whole system there. What are what responses are you getting? What are you seeing? What's surprising, useful, interesting about this prompt? What kind of answers are you getting? You can type in the chat, which I love. Feel free to throw an emoji though if, you know, it's working well, it's giving you good useful information, giving you a good picture. Let me know what you're seeing. I'm guessing everyone's still thinking. This is a fairly large prompt, but it is a personal favorite because it forces the MCP to synthesize across your entire program. So, I am curious to hear, you know, how many of you are getting something on the list maybe you weren't expecting or how you could see this being useful or interesting to use in your day to day or week to week. We'll just wait for a couple responses, and then we'll move on to our next part. Awesome. Thanks for sharing. That's really helpful. It is helpful to know, you know, what to review, what to go work on. Appreciate that. How can someone miss the llama videos? Those are so entertaining. Well, send them a a llama emoji in Slack. You could even have you can prompt your AI tool to do that. Just be like, hey. Send them all a llama. Get them moving. That's great. Okay. Thanks so much for sharing. Really appreciate it. So let's keep us moving just to for the sake of time. But really curious as we keep going, please keep sharing in the chat what's coming up for you. What are you seeing? It's really helpful for for me and the team to to get your answers. So here were three prompts, maybe four, depending on what you did. Three completely different angles on your compliance program, all in plain language. That's the data model you now have access to, and, really, we've barely scratched the surface. Now I think we should try to build something with it. These are just very straightforward analysis, read level prompts. Right? But let's move into build. To set the stage for the build, I want you to notice that everything we just did lived inside Vanta, which is extraordinarily useful. Right? But it's only part of the story. The same way I posted to Slack a few minutes ago during my opening prompt, you can chain any of these queries into action somewhere else. Maybe a ticket in Jira, a message to your team, a pull request in your code base. The answer becomes one stop in a much longer chain, and that's exactly what we're gonna try to do here with our build sprint. So let's go build. This is the main event. You have three build tracks to choose from. This is really where the hands on labeling behind this workshop comes into play. You are going to pick the one that fits where you are today. All three are valuable and all three will produce something real by the end. You can get to the activity cards by navigating to the docs tab there that you see in your screen, and you'll see where it says MCP builder workshop build tracks. Go ahead and click on that to open the doc so you can get to the different tracks you have to pick from, and I will show here what that's gonna look like. Okay. Let me zoom in a little here. Alright. So that should open this doc for you, and it gives you the three tracks you have to pick from. I would say explorer is our more beginner track. Builder would be somewhere around intermediate to more advanced, and then architect is probably gonna feel a bit more advanced. If you click on which track you wanna try, it will take you directly to that section where it will give you directions on what to do. It'll give you the steps. Also, feel free to scroll through if you want to see what the options are so you kinda know what you're getting into before you pick which track. The architect one, again, is going to be that kind of like Vanta plus other tool combo. So if you're really interested in that, then you may wanna jump to architect. So this is the next portion. I'm gonna probably give us around probably eight ish minutes or so to do this portion of the workshop. I'm not gonna be narrating since everyone's gonna be building on their own. But as you build, I want you to feel free to ask questions in the chat, maybe post observations, share what you're working on or what you're building so I can keep reacting and sharing as things come in. Probably also, I'll unmute here and there and jump to my own Claude so you can see how I did some of these prompts. But please make sure you open that doc. That way you can get to the activity sheet, and I will unmute again soon and check-in. Alright. Build something. Let's come back together shortly. Alright. Just checking in. Give me an emoji, maybe something in the chat. Is everyone actively working on their bill track? Brian, I'm doing some digging for you. I will get back to you shortly. Okay. I see some emojis. Great. Perfect. Yes. Keep working. You have time. This isn't a stop yet. Just wanted to check-in, see how everyone's doing. Okay. Great. Keep working. Alright. Probably gonna give another three or four minutes. While you're still working, keep working on your activity. I'm just gonna pop over to my blog to show you what I was working on. So for my explorer, that's kind of the beginner track. Right? I wanted it to generate a compliance status report organized by framework and format so I could share it with my team today. So it's telling me it pulled the live data and filtered out some kind of, like, the sandbox stuff that I have going on in there. And looky there. Pulled out a whole compliance status report for me with an executive summary right here at the top. Told me my frameworks that are in active audit, privacy data protection, security frameworks, and controls catalogs, all sorts of really helpful info. There's key takeaways for the team at the bottom of this report. Now this is where it gets fun. Right? Because now you have this report, but maybe you wanted to have a different tone or you have wanted to have additional context. Maybe you want to use an MCP connector to build something else here, which could be a Slack message or an email. Whatever the case might be, you can now refine this and continue the conversation in order to make sure you have the report that you need, for for this step. So that's how it worked out for me on the explorer track. And I'm gonna let you all keep working. Again, feel free to ask questions and chat if something comes up. We're gonna work for another couple minutes, and, I will show my builder here in one second. Also, Brian, to answer your question, yes. It does look like you can do this. It won't be like an automatic requirement to control mapping situation, but, it will give you kind of, like, an approximation to help you gauge scope. So the definitive gap list might be slightly different, but you can do this to get a sense and an idea of it. So I'm unable to re read the replies to the threads. Someone else may have already said that, but just wanted to make sure you knew. Okay. Let me jump into my builder. So for this one, I wanted it to pick my highest priority failing compliance test, give me the full details, and it did that. Right? The second part would be for you to actually take the action, which could be to, potentially do a fix, which is a bit more intense as we know, or to create a risk entry. So that was sort of that second right action that you could take based on that prompt. And then the third one I didn't do because I'm I modeled that in the beginning, and I wanna know what you all have built. So let me give it one more second, and then we're gonna come together. And I wanna hear what everyone did for their for their builder track. Alright, folks. Let's go ahead and come back together. So if you are, in the weeds in your AI tool, press pause, come back to it. Whatever you have right now is going to be great. The main thing today was kind of showing up, getting a chance to see what type of prompting you could use with our MCP, what type of capability exists there, and how you can make it work within your tool ecosystem. So no matter what you showed up, you connected, and you hopefully built something real, during our last ten minute activity. But I would really love for everyone to drop one line in the chat. What track did you choose? What did you build? And maybe one way you're gonna use it or extend it after this session. So which track, what did you build, and maybe one way you plan to use it or extend it after this session? Everyone must still be very busy building. Okay. Thanks, Chris. Perfect. Oh, I love that. That's so cool. I wish we could screen share because I would love for, peep for us to be able to see what you did, but that's great. Would love to know what everyone else chose or what you're you're maybe still working on or gonna keep trying, after the session ends. And, of course, now you have this document. So if you did explore today, maybe in a couple weeks if you're feeling more comfortable, you could try the next track. And then maybe when you get a little more confident after that, you'll want to try out the architect and see how your tools can work together. Chris is technically kind of already doing that. You're doing an architect because you're you're pushing it out into that other tool, which is amazing. I'm really hoping for one more share. Alright. Maybe I have to give up my dreams. Alright. Hopefully, you all chose a track and built something, or you're going to try to build even more after the session or later this week at some point. And we would love to know what you build, what you create with the MCP. This is really just scratching the surface. But, hopefully, throughout today, you have gotten a little bit more comfortable with testing things out. You know, hopefully, what we've accomplished here in under an hour is that you are connected to Vanta in your AI tool, and you have started to build some sort of real compliance workflow that's best for you. And that's not really a demo. Right? Like, you've actually spent this time not in a webinar, but building something that you can use, test out, refine, as your program does the same. Really excited about the engagement, and we're excited to see what else you can do with it. So let's just move into a little bit of wrap up, a few things to keep the momentum going after today. So the prompt library that I showed has everything you use today plus a lot more. That lives in the help center along with a lot of other cool articles on our MCP. If you wanna go deeper on the engineering side, the cloud code plug in lets you generate infrastructure as code fixes and open pull requests directly from your terminal. So there's some really powerful stuff there that's available to you that is also in our help center documentation and our developer docs. Right actions are expanding fast, so we'll keep the docs updated as new capabilities ship. But as I was prepping for this and spoke with our product manager, she really wanted you to flag gaps. So if you notice gaps, please let your CSM know they can pass along your feedback to the product team. We're always looking to continue expand and meet the needs of our customers. Also, just a call out if you're interested in our CLI, you can also talk to your CSM and get added to the early access for that. The full launch is coming very soon though, so you can always, just wait a little bit here and it will be out in the wild. Also, if you have any questions right now before we leave, please feel free to drop those in the chat or you can always, reach out to us, reach out to support. Before we leave, I wanna make sure you all know about the Vanta community. It's our Slack powered space for security, GRC, and compliance practitioners worldwide. It's one of the best places you could keep learning after today. So people are doing really creative things with the MCP and then sharing about it in the community. It goes way beyond what we've covered in today's exercises. For example, one customer runs cloud code against their failing test every single day and tracks every fix in a Notion dashboard. Another one was giving, was pulling their entire compliance program into an Excel snapshot for an independent backup. A third customer talked about auto generating their subprocessor list straight from Vanta data and so many more. If you want more ideas like that, you can join the community and check out the AI channel. It's full of these. You can join at community.vanta.com or you can scan the QR code that's on this screen, and any Vanta customer is welcome to join. And if you loved our session today, found it helpful or inspiring or even just got you motivated to try out more prompts with Vanta's MCP, we would love for you to tell people about it. Please post about your experience on LinkedIn. You can tag Vanta. You can tag me. You can use our hashtag here on the screen, build with Vanta. I've even included a sample post here that you can use if that's easier. I know sometimes I look at LinkedIn, I'm like, I have no idea what to write here. If that's sometimes you, feel free to take a picture of this, copy paste it, throw it into LinkedIn, and tag us. You can also scan this QR code to follow Vanta if you are not already. And lastly, thanks so much for building with me today. This is our first session. We're gonna continue to refine and update as we go, but this was a really big VIP sneak peek, and we're so happy you were here. This is just the beginning. So much more to come with our MCP and building. So please stay connected, and we hope you have a great rest of the week. Also, if you click on the docs tab, there is a way to give me some specific feedback. I would love to hear how the session was for you and to continue to refine the time. So thank you so much. Have a great week, everyone.