A 3-step AI coding workflow for solo founders | Ryan Carson (5x founder)

1. 0:00 – 3:25

   Introduction and Ryan’s recent AI projects

   1. RCRyan Carson0:00

      I think the biggest mistake that I do, that everyone does, is they try to rush through the context, where you just don't have the patience to tell the AI what it actually needs to know to solve your problem. And I think if we all just slow down a tiny bit and do these two steps, it speeds everything up. Nobody really knows how to do this stuff. The only way you're really gonna figure it out is by getting in here and getting your hands dirty and see what works.

   2. CVClaire Vo0:22

      That's a place where so many engineers and product managers get stuck in a loop. Like, who's gonna take this PRD and actually break it down in the right steps? So even just this is such a time-saver for people building products.

   3. RCRyan Carson0:35

      Building this new startup, I literally feel like I'm able to do all of it. Am I able to do it as well as a dedicated product manager? No. Am I able to think as deeply as a CTO? No. But I am able, for sure, to build this company.

   4. CVClaire Vo0:49

      This is the way, people. I'm telling you, pay attention. [upbeat music] Welcome to How I AI. I'm Claire, product leader and AI obsessive, here on a mission to help you build better with these new tools. 2025 is definitely the year of the vibe coder, but you can't always vibe your way to a scalable execution strategy. In this episode, Ryan Carson, a five-time founder with 20 years experience, shows us how he brings PRDs, task lists, and some advanced prompting techniques to Cursor to make sure he's not just vibing, he's building the right things. Let's get to it. Today's episode is brought to you by ChatPRD. I know that many of you are tuning in to How I AI to learn practical ways you can apply AI and make it easier to build. That's exactly why I built ChatPRD. ChatPRD is an AI copilot that helps you write great product docs, automate tedious coordination work, and get strategic coaching from an expert AI CPO. And it's loved by everyone, from the fastest-growing AI startups to large enterprises with hundreds of PMs. Whether you're trying to vibe code a prototype, teach a first-time PM the ropes, or scale efficiently in a large organization, ChatPRD helps you do better work fast. And we're integrated with the tools you love: v0.dev, Google Drive, Slack, Linear, Confluence, and more, so you don't have to change your workflow to accelerate with AI. Try ChatPRD free at chatprd.ai/howiai, and let's make product fun again. Hey, Ryan, it's nice to have you here!

   5. RCRyan Carson2:26

      Thanks. It's exciting to be here. I've listened to every episode so far, and I'm honored to be here myself. I can't wait.

   6. CVClaire Vo2:32

      So I'm gonna start with an easy question, which is: What are the last three things you built with AI?

   7. RCRyan Carson2:37

      I don't know if you call constantly using ChatGPT with your kids building something in AI, [chuckles] but I feel like, I feel like I'm the constant AI coach in our family, and I'm always delighted, actually, with what our kids are doing. And because of that, uh, my amazing 14-year-old kiddo, Devin, he said, "Dad, like, you know, I've been thinking about this game..." And I said, "Well, like, let's build it." And so we're building a, a primitive little side-scroller, and he's, like, the creative director, so, uh-

   8. CVClaire Vo3:04

      Oh, that's so fun.

   9. RCRyan Carson3:05

      So that's sort of thing one and thing two.

   10. CVClaire Vo3:07

       Mm-hmm.

   11. RCRyan Carson3:07

       I would say thing three is the startup I'm building, which is just this huge amount of work where I'm coding, but then there's just all sorts of little quick things I vibe code, you know, all the time, so, uh, all day, every day.

   12. CVClaire Vo3:19

       All day, every day, ChatGPT-

   13. RCRyan Carson3:21

       [laughs]

   14. CVClaire Vo3:21

       ... and vibe coding. I'm gonna make us all T-shirts and hats-

   15. RCRyan Carson3:24

       [chuckles] Wow-

   16. CVClaire Vo3:24

       ... that say

2. 3:25 – 5:00

   Demo: Creating a PRD with Cursor

   1. CVClaire Vo3:25

      that

   2. RCRyan Carson3:25

      I'd buy that T-shirt. [chuckles]

   3. CVClaire Vo3:26

      Okay, so one of the things that I think you do really well compared to other vibe coders is you bring some structure to the process. You're, you're quite wise about how you use Cursor, and so I'd love for you to pull up your screen and show us how you get Cursor to follow a plan.

   4. RCRyan Carson3:43

      Let's do it. So the reason why, uh, I do this is because I've been coding, coding, coding with AI, and you just learn as you do this that you really have to get good about context, wh- what you're showing the AI, what you're asking the AI to do, and you end up cutting back, uh, all, uh, all the things you're doing to a manageable amount of things, uh, that it- the AI can actually do. So the process I'm gonna show you... I'm in Cursor now. Uh, in case you're listening to the show, you can download it for free at cursor.com. Uh, it's basically a VS Code fork, right? So if you ever use VS Code, it's a great tool. So what I've done here is opened up a, a basic project that I vibe coded yesterday, just to kinda show you w- what's going on. This is [chuckles] a, a stupid little CRM tool for a yacht club, 'cause I thought that would be kind of fun.

   5. CVClaire Vo4:33

      [laughs]

   6. RCRyan Carson4:34

      [chuckles] So, so there's that. So all right, so say that you wanna make a change in here, and it's a bigger change than just a small, uh, you know, a, a quick, "Hey, will you change this thing?" Say it's larger. All right, so you would probably wanna create a product requirement doc, right, like PRD, and, and if you're watching the show and you don't know about Claire's ChatPRD-

   7. CVClaire Vo4:52

      [laughs]

   8. RCRyan Carson4:53

      ... absolutely check it out, 'cause it's amazing. But if you need a lighter lift, then here's what I do. So I've created three

3. 5:00 – 9:53

   Ryan’s open source links: https://github.com/snarktank/ai-dev-tasks

   1. RCRyan Carson5:00

      files over here in my Cursor Rules folder, and I'll walk you through what they do and how they work. I've also open-sourced these, and so we can throw those, uh, a link in the show notes for you-

   2. CVClaire Vo5:12

      Great

   3. RCRyan Carson5:12

      ... all to grab that. So the first one is a simple, uh, instruction for the AI to create a product requirement docs. Uh, again, Claire knows everything about this and-

   4. CVClaire Vo5:21

      Mm-hmm

   5. RCRyan Carson5:22

      ... lives this and breathes this, but to a lot of people that don't know about PRDs, right, it's, it's how you describe a feature that you're wanting to build. And so this rule explains to the AI how to write a PRD for the user. So the way you use this is, is very simple, right? So I'm gonna go over-

   6. CVClaire Vo5:40

      And I wanna pause really quickly because I love your initial prompt, which says-

   7. RCRyan Carson5:45

      Indeed. [chuckles]

   8. CVClaire Vo5:45

      ... "This is a PRD that's suitable for a junior developer to understand and implement this feature." That's such an interesting little call-out.

   9. RCRyan Carson5:53

      Yeah, that... Well-spotted-

   10. CVClaire Vo5:54

       Mm-hmm

   11. RCRyan Carson5:54

       ... because, again, the, as you code, and code, and code with AIs, you, you start to realize that-... They're like a genius PhD student, right? But, but they, they can't seem to connect these really s- simple, obvious things that you and I know. And so saying junior developer is kind of a way to instruct the AI, let's keep this at the, at a certain level.

   12. CVClaire Vo6:18

       Yep.

   13. RCRyan Carson6:19

       You know, we've got some process, we've got some clarifying questions. Uh, you'll see this in action. I'll actually run it.

   14. CVClaire Vo6:24

       Great.

   15. RCRyan Carson6:24

       So let's do that.

   16. CVClaire Vo6:25

       Great.

   17. RCRyan Carson6:26

       So I'm gonna @include this file, which puts it in the context window, and then I'm gonna give it a simple instruction, which I've pre-written, which is, "I'd like to add a report that shows me all the boat names of members and how many emails they've been sent." This is a big deal if you're in a yacht, in a yacht club. [chuckles]

   18. CVClaire Vo6:42

       You have a lot more fun use cases than [chuckles] I do for vibe coding.

   19. RCRyan Carson6:47

       I, I try, you know. So let's see what happens. So I'm gonna, uh, go ahead and hit Go. We're in agent mode. For those of you who are astute, you'll notice I'm using Claude, uh, three seven Sonnet in max mode. I actually tend to use, uh, Gemini 2.5 Pro, and it's funny, I didn't notice that I had selected that, but fine.

   20. CVClaire Vo7:08

       I'm, I'm now a default o3 girl.

   21. RCRyan Carson7:10

       Ooh! Ooh, we'll have to talk.

   22. CVClaire Vo7:11

       Yeah, and then when o3 gets stalled out, I go to three seven Sonnet.

   23. RCRyan Carson7:14

       Right, right. It's funny, I'm pretty much default, uh, Gemini 2.5 Pro.

   24. CVClaire Vo7:18

       Mm-hmm.

   25. RCRyan Carson7:19

       I love the max mode. It is sort of expensive. I probably spend maybe-

   26. CVClaire Vo7:23

       Yeah

   27. RCRyan Carson7:23

       ... three, 400 bucks a month. Uh, uh, but, you know.

   28. CVClaire Vo7:27

       Worth it.

   29. RCRyan Carson7:27

       I think that's okay. Okay, so what we've got here is the AI came back and gave us some clarifying questions on the PRD. Uh, I'm gonna answer a couple of these just to, to show you how this works. So if you're listening along, the AI has said, "Okay, great, I'll help you create a PRD. Uh, I have a couple questions for you. Uh, the first one is: 'What is the problem, uh, this report is, is trying to solve?'" Well, we are... Uh, what problem is this? We're trying to give visibility into how many emails people are getting, right? All right, so that's thing one. Let's go- let's answer the second one. "Who specifically will be using the report?" All right, we'll say admins. "Where should this report be accessible?" Uh, you pick. And so there's a couple... I, I would say there's probably too many questions in here, but one of these things, again, that I want everyone watching to kind of get used to is, you'll notice in this PRD rule file, I've done something specific here, where I've said, "I want these questions to be dot notation, where it's 2.1 and 2.2," because you end up... Otherwise, the AI would just give you a bunch of questions, and it would put, it would put more than one question in a bullet point, and it becomes hard to use. So you kind of get used to this specificity, right? So I'm just gonna, uh... I'm gonna say the rest, make your best judgment.

   30. CVClaire Vo8:41

       I do the exact same prompt. I say, "You pick whatever you think is best." [chuckles]
4. 9:53 – 11:00

   Quick recap and common mistakes to avoid

   1. RCRyan Carson9:53

      into Cursor.

   2. CVClaire Vo9:54

      Just to kind of recap where we are so far. So you have a Cursor rules files that gives, uh, rules and instructions for generating a PRD. You generate the PRD in agent mode, using whatever your preferred model is. That puts that PRD markdown file in a tasks folder that you've put in, in your repo, and now we're gonna look at that PRD and show how you work through it to build something.

   3. RCRyan Carson10:20

      Exactly.

   4. CVClaire Vo10:21

      Okay.

   5. RCRyan Carson10:22

      You're a good student. So here is our PRD. It's pretty straightforward if you've ever looked at a PRD. Um, I'm gonna show you. So you've got functional requirements, uh, non-goals, design considerations, et cetera, right? And let's back up. Like, what are we doing here? What we're doing is making it clear, uh, to the large language model what we want done. I think the biggest mistake that I do, that everyone does, is they try to rush through the context, where you just don't have the patience to tell the AI what it actually needs to know to solve your problem. And I think if we all just slow down a tiny bit and do these two steps, it speeds

5. 11:00 – 15:31

   Demo: Generating a task list from the PRD

   1. RCRyan Carson11:00

      everything up, right? So we've got a PRD. It's not rocket science here. Uh, now, next, what we wanna do is generate tasks for this. So let's go and generate the tasks based o- off this PRD. So, I... If you're listening, I'm back in Cursor, and I'm, uh, I'm including the, the file that generates the tasks. It's called Generate Tasks, and I'm gonna say, uh, "Please generate the tasks." Tasks for, and then I'm gonna tag the PRD.

   2. CVClaire Vo11:28

      Got it.

   3. RCRyan Carson11:29

      Do, do, do.

   4. CVClaire Vo11:29

      So this is another rule, probably similar to your Generate PRD, one that explains what a task is and how to do it well, and then you're giving it the context of the PRD itself.

   5. RCRyan Carson11:38

      Exactly. So let's have a quick look at the generate task list. So the goal is to guide an AI assistant in creating a detailed step-by-step task list, list. This is what it- what I want the list to look like. This is the process. You'll see, it's gonna ask me a couple questions, and then it's gonna pump out, uh, tasks in this format as a markdown file.

   6. CVClaire Vo11:58

      I have to ask, how did you create these-... rules. These, the, it's a lovely prompt. It's well-structured, it's clear. How did you get good at writing these instructions prompts?

   7. RCRyan Carson12:09

      The same way you did. Basically, I tried a couple things-

   8. CVClaire Vo12:13

      Yeah

   9. RCRyan Carson12:13

      ... they didn't work. I got more specific, and then, of course, you know, I had a very intelligent LLM write this for me.

   10. CVClaire Vo12:20

       Yep.

   11. RCRyan Carson12:20

       And then I edited it. You know, you- again, you sort of learn some of the tricks of the trade. Like, I want this task list to be in Markdown, and I want, I want there to be check boxes so we can check them off. I mean, [chuckles] silly stuff like that. So the biggest thing I want people to walk away from this show is, you know what? Nobody really knows how to do this stuff. The only way you're really gonna figure it out is by getting in here and getting your hands dirty, and see what works. Um, and then stick with a model that you consistently like. Like, I'm getting to know Gemini 2.5 Pro really well, like, what it's good at, what it's not good at. So let's see what it's good at. All right, so I just said, "Please generate tasks for this PRD." I tagged the file. Uh, we're using a reasoning model, so we're seeing, uh, the thinking tokens kind of whizz by.

   12. CVClaire Vo13:09

       And all this thinking might cost you a little bit more, but you get a little bit more visibility, you learn more, and looks like it's doing a good job.

   13. RCRyan Carson13:16

       Amen. Yeah, I f- I feel like for the extra five cents, like, this is absolutely worth it. Um, so in, in this instruction, it, it... The instruction is, "Give me some basic tasks, and then ask me if they're okay, and then tell me to go to proceed." So you'll see, it says, "Ready to generate the subtasks? Uh, respond with Go to proceed."

   14. CVClaire Vo13:38

       You know, as I'm looking at this side by side, what I really like about your generate tasks instructions is you've given it in a very explicit process. It's you get this file, you do step one, you get this next step f- uh, next file. And it's not quite an agent, but it really brings in this, like, agentic thinking of where are the decision points, where are the user interaction points, um, but in sort of more this linear, um, step-by-step chat mode.

   15. RCRyan Carson14:05

       Exactly. And I have been hounding the Cursor team [chuckles] non-stop. Like, "Why don't you just build this into the core developer experience of the app? I, I don't understand why this isn't just the way you use Cursor." And they keep saying, "We're doing it," so you know.

   16. CVClaire Vo14:18

       And, and for those listening, that delightful ba-doo-

   17. RCRyan Carson14:21

       Uh-huh

   18. CVClaire Vo14:21

       ... is when Cursor [chuckles] Cursor is done generating.

   19. RCRyan Carson14:24

       Isn't that the best? I love that.

   20. CVClaire Vo14:25

       I love it. This episode is brought to you by Notion. Notion is now your do-everything AI tool for work. With new AI meeting notes, Enterprise Search, and Research Mode, everyone on your team gets a note-taker, researcher, doc drafter, brainstormer. Your new AI team is here, right where your team already works. I've been a longtime Notion user, and have been using the new Notion AI features for the last few weeks. I can't imagine working without them. AI meeting notes are a game changer. The summaries are accurate, and extracting action items is super useful. For stand-ups, team meetings, one-on-ones, customer interviews, and, yes, podcast prep, Notion's AI meeting notes are now an essential part of my team's workflow. The fastest-growing companies like OpenAI, Ramp, Vercel, and Cursor all use Notion to get more done. Try all of Notion's new AI features for free by signing up with your work email at notion.com/howiai.

6. 15:31 – 18:07

   The importance of context when working with LLMs

   1. RCRyan Carson15:32

      So we're gonna accept those changes. We're gonna hop over here, and let's just kinda see what it did, right? So, uh-

   2. CVClaire Vo15:37

      Oh, wow!

   3. RCRyan Carson15:38

      I know, it's fun, isn't it? So what you've got up here is the relevant files. Now, this is a trick I learned from a friend on X, where my thinking here is that this should help the LLM just remember what files are we really focusing on. Even though I will specifically tag these in the context, uh, I, I think it's helpful. One of the things I want, uh, to briefly touch in, I probably anthropomorphize LLMs-

   4. CVClaire Vo16:02

      Mm-hmm

   5. RCRyan Carson16:02

      ... probably too much.

   6. CVClaire Vo16:04

      [chuckles]

   7. RCRyan Carson16:04

      But, but because they're trained on human output on the web, you know, my belief is, th- we need to give LLMs, like, the right context and be as helpful as we can so they can actually solve our problems, right?

   8. CVClaire Vo16:17

      I, I completely agree. I'm also very polite to-

   9. RCRyan Carson16:20

      Yep, me too

   10. CVClaire Vo16:20

       ... to the LLMs. It's how I get people to, people to do work. [chuckles]

   11. RCRyan Carson16:24

       Yeah.

   12. CVClaire Vo16:24

       It's how I'm gonna get the agents to do work.

   13. RCRyan Carson16:25

       Right, so why wouldn't you, you know, be- treat an agent like you would treat a human? I- that's the way I look at it.

   14. CVClaire Vo16:30

       Yeah.

   15. RCRyan Carson16:30

       So-

   16. CVClaire Vo16:30

       I agree.

   17. RCRyan Carson16:31

       All right, so then we've got, you know, pretty detailed list of tasks, right? We've got one, two, three, four with subtasks. Some we've even got sub-subtasks here, and then we're gonna start iterating on this. So here's how it works. I've got another rule called Task List, um, which I probably need to re- rename, 'cause that doesn't really make sense. But this is the instructions for iterating through, uh, these tasks, so I'll kinda walk you through that. So this is task list management. These are guidelines for managing task lists in Markdown for files to track progress. The task completion, we have a, uh... We wanna do one subtask at a time. Like, this is really important, that the, the AI doesn't start trying to do all the tasks. Uh, when you finish a subtask, immediately mark it as complete, and then you'll see I say, "Stop after each subtask," uh, and, "Wait for the user's go ahead." And there's just a little bit more, uh, clear- clarity here.

   18. CVClaire Vo17:26

       If people see me shaking my head-

   19. RCRyan Carson17:28

       [chuckles]

   20. CVClaire Vo17:28

       ... it's because I am realizing now I'm using Cursor like such an amateur! This is- [chuckles]

   21. RCRyan Carson17:34

       I doubt that.

   22. CVClaire Vo17:34

       This is so good. I'm just floating through the ether saying-

   23. RCRyan Carson17:38

       Just, "Woo-hoo!"

   24. CVClaire Vo17:39

       ... you know, "o3, take me away."

   25. RCRyan Carson17:41

       Solve all my problems.

   26. CVClaire Vo17:41

       And maybe I'm, I'm overconfident in my product management skills, and so I'm putting this all into the chat. But this is, this is the way, people. I'm telling you, pay attention.

   27. RCRyan Carson17:52

       I appreciate that. I mean, I think it- the answer is yes to both those. I mean, sometimes you do just need to roll and see what happens, but I'm learning over and over that if I don't follow this process, I end up down some rabbit hole, and I have to revert, right?... So let's,

7. 18:07 – 18:56

   Demo: Working through tasks systematically using Cursor

   1. RCRyan Carson18:07

      let's go ahead and continue here. I'm- what I'm doing now is I'm tagging the task list rule, which tells the AI how to act, and then I'm gonna say, "Let's start," and then I'm gonna tag the, uh, the tasks.

   2. CVClaire Vo18:18

      The task list.

   3. RCRyan Carson18:20

      Yeah, which is, uh-

   4. CVClaire Vo18:21

      So we got a PRD, we got a task list, and now we have a set of rules that knows how to work through a task list and actually get work done.

   5. RCRyan Carson18:30

      It does, exactly.

   6. CVClaire Vo18:31

      Okay.

   7. RCRyan Carson18:31

      So, all right, so let's start. It's gonna think about it, and it's looking through this big list of tasks right now. And so what it's saying is, "Okay, let's start with the first subtask: Define Prisma schema email campaign. I'll start by reading the existing Prisma, uh, campaign. Okay, I see existing blah, blah, blah." You know, so it's just thinking it through, and boom, [device notification] it has checked off 1.1.

8. 18:56 – 20:00

   Change management

   1. CVClaire Vo18:56

      Ah, with a delightful noise!

   2. RCRyan Carson18:57

      I know.

   3. CVClaire Vo18:58

      I have a question. Are you, are you hooking it all into Git? How are you doing sort of-

   4. RCRyan Carson19:04

      Mm-hmm

   5. CVClaire Vo19:05

      ... the, the change management here?

   6. RCRyan Carson19:07

      Yep, so what I tend to do is I will commit after I, after I finish either one of the bullet-

   7. CVClaire Vo19:13

      Yeah

   8. RCRyan Carson19:13

      ... one of the, one of the parent tasks.

   9. CVClaire Vo19:15

      Yep.

   10. RCRyan Carson19:15

       If I, if I feel like the app is in a state that it's workable-

   11. CVClaire Vo19:19

       Yep

   12. RCRyan Carson19:19

       ... then I will commit at that point.

   13. CVClaire Vo19:21

       Got it.

   14. RCRyan Carson19:21

       If I don't, I won't commit until all of these tasks are done.

   15. CVClaire Vo19:26

       Oh, wow, okay.

   16. RCRyan Carson19:26

       Um, whi- which, you know, is probably, you know, half a day, uh, uh, of work. It you, you know, you kinda get used to, like, "Oh, all right, if I had to revert now-

   17. CVClaire Vo19:37

       Yeah

   18. RCRyan Carson19:38

       ... you know, how bad would it be?"

   19. CVClaire Vo19:39

       Yep.

   20. RCRyan Carson19:39

       And, and I try to think about, "Well, what would I need to undo?" So it said, uh, "All right, shall I proceed with subtask 1.2?" And I'm gonna say, uh, "Yes." Sometimes I'm really lazy, and I just say, "Y" [chuckles] a- as in the letter Y. So it's just gonna keep working. Now, we don't need to go through this whole

9. 20:00 – 21:50

   How task lists save time for product managers

   1. RCRyan Carson20:00

      thing.

   2. CVClaire Vo20:00

      Yeah, yeah.

   3. RCRyan Carson20:01

      Um, but I just wanted to show, uh, all of your amazing listeners and audience, [device notification] this is a pretty easy process to follow, and I've built huge features w- with this. I mean, you know, 10,000 lines of code reliably and almost never had trouble. I still feel like this human-in-the-loop part is really important, where after each task, you are kind of checking what's happening. I've noticed that it often does introduce some small problem, or there's a linter-

   4. CVClaire Vo20:31

      Yep

   5. RCRyan Carson20:31

      ... um, you know, error, uh, and then you gotta go fix it, so, uh-

   6. CVClaire Vo20:35

      And, you know, this is great to get the actual engineering work done, but if I just take a step back, for product managers out there that don't know where to get started with Cursor, even if you just did the PRD'd task list part... I'm looking at this task list right now, and it's got, and basically, like, epics and tasks in it, and that's a place where so many engineers and product managers get stuck in a loop. Like, who's gonna take this PRD and actually break it down in the right steps that are gonna work in our code base? So even just this is such a time-saver for people building products.

   7. RCRyan Carson21:10

      I think so. And now, you can over-engineer this process.

   8. CVClaire Vo21:14

      Mm-hmm.

   9. RCRyan Carson21:15

      You know, this is literally a markdown file.

   10. CVClaire Vo21:17

       Yep.

   11. RCRyan Carson21:17

       Um, it's somewhat hand-cranked. You know, I thought, "Oh, maybe I'll use, you know, Asana's MCP, you know-

   12. CVClaire Vo21:23

       Yeah, yeah

   13. RCRyan Carson21:23

       ... server and create Asana tasks," and-

   14. CVClaire Vo21:25

       Yeah

   15. RCRyan Carson21:26

       ... and I was like, "No." Like, it's actually easier for me just to see a markdown file-

   16. CVClaire Vo21:29

       Yeah

   17. RCRyan Carson21:29

       ... and know what's happening, and I can even add tasks to it. So my encouragement to everybody is just start small, start simple, and get good at that and get comfortable with that, and then you can, you know, uh, graduate from there, so.

   18. CVClaire Vo21:43

       And, and 30 chime noises later, we are going to have a report about your yacht club-

   19. RCRyan Carson21:49

       That's right

   20. CVClaire Vo21:49

       ... emails.

10. 21:50 – 24:50

    Demo: Using MCPs for front-end testing

    1. CVClaire Vo21:50

       I mean, speaking of MCPs, are you using MCPs at all in your Cursor experience? How is that fitting into your workflow?

    2. RCRyan Carson21:56

       I am. So I'm gonna, um, show you a couple of those now. Um, I wasn't using many at the beginning, and then the first MCP I started using was for Postgres.

    3. CVClaire Vo22:07

       Mm-hmm.

    4. RCRyan Carson22:07

       Um, because it's really useful to ask the AI, "Hey, can you go see if I, if this data is actually in the database?" Um, so I, I started that, and then I've gone down the rabbit hole. So I'm gonna show you a couple, uh, MCP servers that I think are really useful, so let me go ahead and share my screen. Okay, so what you are seeing now is, on the left, I've got Browserbase.

    5. CVClaire Vo22:27

       Mm.

    6. RCRyan Carson22:27

       It's the actual backend. I've got a free account, so I just wanna try this out and see how well it works. I'm gonna show you, uh, the fun that you can do with an MCP here. So on the right, I've got Cursor, and I'm in my Cursor settings, and I go to MCP.

    7. CVClaire Vo22:41

       Mm-hmm.

    8. RCRyan Carson22:41

       And you'll see down here, I've got a Browserbase MCP-

    9. CVClaire Vo22:45

       Yep

    10. RCRyan Carson22:46

        ... which I've set up. I've got Stagehand as well, which is really fun. So let me show you, uh, how this works. So we're gonna go back into... I'm gonna start a new chat. Turn on agent mode, and I'm gonna say, um, "Navigate to ChatPRD," yeah, and take a screenshot. Now, let's see if the MCP gods, uh, cooperate. Because I did this this morning, and it-

    11. CVClaire Vo23:15

        Mm-hmm

    12. RCRyan Carson23:16

        ... I w- I was like, "Wow, [chuckles] this, this is super cool!" Okay, so what we should see over here in a minute... Let me refresh this to see. It did. Oh, my God, this is so cool. Okay, so in the, in the cloud-

    13. CVClaire Vo23:32

        Ooh!

    14. RCRyan Carson23:34

        So what I'm doing is controlling a headless browser in the cloud from Cursor. [chuckles] What?

    15. CVClaire Vo23:41

        The future is now. [chuckles]

    16. RCRyan Carson23:43

        It's in... It's bonkers. So, well, it's like, okay, what use is this?

    17. CVClaire Vo23:47

        Uh-huh.

    18. RCRyan Carson23:47

        Like, that's kind of cool. I mean, let's actually do something kind of fun. So I'll say, uh, "Navigate to pricing." Okay, so we should actually see the cursor move over here in this headless browser in the cloud.

    19. CVClaire Vo24:04

        You're just showing off at this point.

    20. RCRyan Carson24:06

        I... Well, I just learned this this morning, and, and-

    21. CVClaire Vo24:08

        Okay

    22. RCRyan Carson24:08

        ... and I was thinking, "Oh, my God, this is so neat." Let's see if it, if it actually works.... Yep, there it goes.

    23. CVClaire Vo24:15

        Ah, whoa!

    24. RCRyan Carson24:17

        So again, well, besides kind of a parlor trick, like what-

    25. CVClaire Vo24:21

        Yeah

    26. RCRyan Carson24:21

        ... actual use is this, I think it's going to unlock a huge amount of front-end testing for me.

    27. CVClaire Vo24:27

        Yeah.

    28. RCRyan Carson24:28

        Right? So a- as you know, like, it's really kludgy right now trying to, uh, to squash bugs on the front end when you're, like, taking a screen grab, and you're pasting into Cursor-

    29. CVClaire Vo24:38

        Yeah

    30. RCRyan Carson24:38

        ... and you're like, "No, I mean over there." And, and-
11. 24:50 – 26:45

    Specific MCPs and what to use them for

    1. RCRyan Carson24:50

       show you. Now, if you're watching this, you're like: What? Uh, it- really what we're doing here is... Uh, I'll go back to the MCP, uh, part of Cursor. What we're doing is, is basically giving Cursor the ability to interact with other apps, right? And so we're saying, "Okay, Browserbase is pretty cool. You can have this browser in the cloud you can do things with." I wanna be able to tell the AI what to do. I don't wanna have to know how to call the APIs and do all that stuff, so that is, uh, what I thought was a lot of fun. So-

    2. CVClaire Vo25:22

       And can you walk us through just a couple of the use cases of the ones we're seeing here? So Browserbase lets you basically browse the web and, and do a couple things through, through the text window in Cursor. What are some of these other ones that you find useful?

    3. RCRyan Carson25:36

       Um, so Postgres is probably the one I use the most.

    4. CVClaire Vo25:38

       Mm.

    5. RCRyan Carson25:38

       Um, so this... For the startup I'm building, I'm using Postgres on, on, uh, for the database in Vercel, and there's just a lot of times where I just wanna be able to tell the AI, you know, "Is this value in this row in the database?" Like, I don't wanna have to actually write-

    6. CVClaire Vo25:54

       Yeah

    7. RCRyan Carson25:54

       ... SQL to do that.

    8. CVClaire Vo25:55

       Yeah.

    9. RCRyan Carson25:56

       And so you can go right into, uh, the chat window and just say, "Hey, you know, u- use the Postgres tool, and tell me, uh, if this is in the database or not," which is really cool. Prisma is... I'm using for this, uh, play project, wh- which is fun, same with SQLite. So the one I use every day is, is Postgres.

    10. CVClaire Vo26:14

        And one of the things that I love about AI, especially in the sort of dev tools stack, is it just reduces toil. And one of the pieces I feel like is, um, toil for engineers is how many tabs you have to have open to orchestrate working across your task list, right? You have, you have to have your project management software open, so you know, like, what task I'm working on and what's next. You have to have your browser open, so you can do some work. You're querying your database, and all this puts it in a single interface that you can seamlessly switch through

12. 26:45 – 31:23

    Demo: Using Repo Prompt to gain precise control over context

    1. CVClaire Vo26:45

       in natural language.

    2. RCRyan Carson26:46

       Amen. So I do wanna show you one tiny, quick other tool, and I know-

    3. CVClaire Vo26:49

       Yes, please

    4. RCRyan Carson26:49

       ... we're almost out of time.

    5. CVClaire Vo26:50

       No, let's do it.

    6. RCRyan Carson26:50

       Um, so you, you're probably hearing over and over, uh, everyone's listening or watching me preaching about context.

    7. CVClaire Vo26:57

       Yep.

    8. RCRyan Carson26:57

       Um, it, it is just so much more important than I think we understand. And again, if we anthropomorphize, it's like, well, how would you expect anybody to do anything unless you give them the right book or the right piece of paper, right? And so I've started to use a tool called Repo Prompt. Again, I don't have any financial, uh, uh, reason to say this. I don't own any of this company, but it's this really great tool for Mac, and I just wanna show you how to use it really quick. The question is why? Like, why wouldn't I just use Cursor? Well, the, the thing about Cursor is there's all this magic happening in the background with the context, where you don't know for sure what is in the context unless you tag it, right? Which is fine, but they, they sort of magic that away, and sometimes you really, really wanna control the context. So what I've done is I've opened Repo Prompt. I've got the, the site on the left, and on the right I've got the UI, and I've opened up the How I AI, uh, project, which is that simple, uh, yacht email app. Okay, so say that I want to throw a lot of these into, uh, a prompt to do something. So if I select the whole repo... Let's go over to Compose. Down here, you can see how many tokens that is. All right, so that's 395,000 tokens. Okay, that's way, that's way too many. So let's go ahead and reduce that. So all right, I know I want some stuff in the app folder. All right, why is it so big? Now we're at, you know, 324,000. Still too big, so I think it's probably in generated, so you get rid of that. Okay, good, now we're at 12,000 tokens. So the point is you- if you know what context you want, Repo Prompt is a really powerful way to do that. So I'm gonna go ahead and select Components, Lib, uh, let's open up the schema for Prisma, you know, and scripts. All right, now what? So I go up here, and I put in a prompt: How can I improve the maintain- maintainability of this code? And then you can do some other neat things, like I wanna include a stored prompt, and I'm gonna use the architect version. Now, what is that? Well, that is a prompt that Repo Prompt has written already with a bunch of kind of power moves in it, right? It's- it should act like a, a- an architect versus a dev. So now I go down to Copy, and I'm gonna say, "I wanna include the saved prompts, I wanna include the files, and I wanna include the user instructions," and, and I'm gonna click Copy. Now what? Well, I'll show you. So let's go over, uh, to o3. Now, I could, I could go over to Cursor, but I don't ex- know exactly what Cursor is gonna do to my context. So instead, I'm gonna paste it in here. Now, what is this? What you've got... Let's start at the bottom. It's basically putting everything in XML tags, right? So you've got user instructions, so that was the prompt I put in. Then the meta prompt is, uh, telling it how to, to execute the prompt and, and the files that it's including. Let's go up here, so file contents. This is kind of the key. So each file is included, and it's, it's demarked specifically like this. So it says the file with the actual path, and it's... And so it basically, it's very specific. It's saying to the AI, "This is exactly the context you need, and it's super clear," and then you can execute some sort of prompt on that.... I do this for, like, heavy lifting stuff, where I used to do this all the time with o1 Pro, where I would go into Repo Prompt, I would select exactly the right context, and then I would go into o1 Pro and say, "Think super hard about this. I've given you exactly the right context," and you get amazing answers 'cause of that.

    9. CVClaire Vo30:40

       Yeah, and these, these new models have such big context windows, but I'm not copying and pasting 12,000 tokens into the chat window.

    10. RCRyan Carson30:49

        Right.

    11. CVClaire Vo30:50

        But this tool just does it for you. Really interesting!

    12. RCRyan Carson30:53

        Oh.

    13. CVClaire Vo30:54

        And then you're getting less of a black box here.

    14. RCRyan Carson30:57

        Less of a black box.

    15. CVClaire Vo30:57

        Mm-hmm.

    16. RCRyan Carson30:58

        Yeah, and, you know, and I have a feeling that this stuff will probably go away. I think, you know, the context windows are gonna get bigger. Tools like Cursor are gonna get better at managing context, you know, but right now you have to do all this stuff. Like, you, you can't just wave your wand and, and hope the LM's gonna write all the right stuff for you.

    17. CVClaire Vo31:17

        That- that's what I've been doing. So now [laughing]

    18. RCRyan Carson31:19

        [laughing]

    19. CVClaire Vo31:19

        ... I have a, now I have a better process-

    20. RCRyan Carson31:22

        Please work

    21. CVClaire Vo31:22

        ... um,

13. 31:23 – 32:10

    Music’s role in Ryan’s development stack

    1. CVClaire Vo31:23

       to, to follow here. Okay, anything else really important in your stack you need to tell us about?

    2. RCRyan Carson31:27

       Um, the most important thing-

    3. CVClaire Vo31:29

       Mm-hmm

    4. RCRyan Carson31:29

       ... um, uh, is this. So let me share my screen. So this is Tiesto's live set from New York City. [laughing] So this is how I actually code. I just turn on some EDM, and-

    5. CVClaire Vo31:40

       Yeah

    6. RCRyan Carson31:41

       ... and late at night, you know, code, uh, after everyone goes to bed, listening to amazing EDM, and, uh, that is an important part of my stack.

    7. CVClaire Vo31:48

       So if I could ask for an additional Cursor feature, it would be AI-generated streaming EDM matched to the generation pace of your tokens that ends with a drop instead of that cursor. [laughing]

    8. RCRyan Carson32:02

       I would pay extra money for that.

    9. CVClaire Vo32:04

       You're- you heard it here first. Get, get... Forget all this task management context window stuff. [laughing]

    10. RCRyan Carson32:08

        Forget about that. I want EDM at the end.

    11. CVClaire Vo32:09

        Let's bring

14. 32:10 – 34:44

    Lightning round and final thoughts

    1. CVClaire Vo32:10

       in the fun. Okay, we'll spend three minutes on lightning round questions. One, you're a builder, you're a founder. It's clear how this technology is changing the building part of things. How is it, in your mind, changing the company and the founder side of things?

    2. RCRyan Carson32:25

       Wow, it's a complete rewrite. So I've been fortunate to start three companies and, and see them acquired, and, and one of them, you know, we had, uh, about 110 employees, and I had a CTO and a VP of engineering and, you know, product managers, and, and, you know, building this new startup, I literally feel like I'm able to do all of it. Now, am I able to do it as well as a dedicated product manager? No. Am I able to think as deeply as a CTO? No. But I am able, for sure, to build this company by myself, and, I mean, you're doing this with ChatPRD. I mean, it's, it's bonkers to me that it's actually possible, so I, I just can't wait for the future.

    3. CVClaire Vo33:08

       I could not agree more. Okay, and then, of course, you showed us how you're a very thoughtful and organized product manager and manager of your AI, but you know it sometimes doesn't listen. So what do you do? What's your, what's your tactic for really getting it back on track?

    4. RCRyan Carson33:28

       So I'm just too nice. I, I just say, "Please think harder about this." [laughing] Like, "I know you can do this. Um, you know, think again about it." And I'm just not a mean person, so as much as I wanna say, you know, "God damn it," you know-

    5. CVClaire Vo33:45

       [laughing]

    6. RCRyan Carson33:46

       ... I don't.

    7. CVClaire Vo33:47

       I, I have a hypothesis that this is, this is actually a parenting thing 'cause I do the same thing. I say, "I believe you can do this."

    8. RCRyan Carson33:55

       [laughing]

    9. CVClaire Vo33:55

       "I believe in you."

    10. RCRyan Carson33:57

        I, I believe you can.

    11. CVClaire Vo33:57

        I believe you can. Well, that's amazing. Ryan, where can we find you, and how can we be helpful?

    12. RCRyan Carson34:02

        Um, I am on X all the time-

    13. CVClaire Vo34:04

        Mm-hmm

    14. RCRyan Carson34:04

        ... so x.com/ryancarson. Um, and if you wanna know a little bit more, just head to ryancarson.com, and that is me.

    15. CVClaire Vo34:12

        Great. Well, thank you so much.

    16. RCRyan Carson34:13

        It's been a blast. Thank you. [upbeat music]

    17. CVClaire Vo34:17

        Thanks so much for watching. If you enjoyed the show, please like and subscribe here on YouTube, or even better, leave us a comment with your thoughts. You can also find this podcast on Apple Podcasts, Spotify, or your favorite podcast app. Please consider leaving us a rating and review, which will help others find the show. You can see all our episodes and learn more about the show at howiaipod.com. See you next time! [upbeat music]

Episode duration: 34:44