Section Insights
Introduction to Agentic Engineering
What is the concept of agentic engineering?
Matt discusses his approach to building AI tools without traditional engineering skills, emphasizing the value of compound engineering.
- Compound engineering is a key tool for building valuable AI applications.
- One does not need to be an engineer to create impactful technology.
- The ability to ship products has evolved significantly with new tools.
The Shift in AI Tool Usability
When did AI tools become effective for real projects?
Matt highlights a pivotal moment around Thanksgiving last year when AI tools transitioned from being toys to valuable resources for building applications.
- The launch of Codex marked a significant improvement in AI tool capabilities.
- There was a clear distinction between pre- and post- Thanksgiving in terms of usability.
- Early adopters faced challenges but saw potential in new AI technologies.
The Launch of Printing Press
What is the printing press and how does it work?
Matt explains the printing press as a tool that allows users to discover and utilize secret APIs, exemplified by a recent change in Domino's API.
- The printing press enables users to create command-line interfaces (CLIs) easily.
- Real-time updates to APIs can impact functionality, as seen with Domino's.
- The community has rapidly expanded the library of available CLIs.
Community Adoption of Printing Press
How is the printing press being adopted by businesses?
Matt shares examples of companies opting to use the printing press for their official CLI instead of building their own solutions.
- Businesses are recognizing the efficiency of using existing tools like the printing press.
- Open-source solutions are gaining traction in the developer community.
- The ease of testing and building with new tools fosters innovation.
Contributions and AI in Development
How does Matt integrate AI in his development process?
Matt describes his experience contributing to projects using AI, including the challenges and interactions with other developers.
- Transparency about using AI in coding is important in collaborative environments.
- AI can assist in coding but requires careful oversight to avoid pitfalls.
- Engagement with the developer community can lead to innovative solutions.
Transcript
0:00 I certainly don't read any code. I don't even know how to read any code. >> I thought you were actually an engineer. >> I'm not an engineer. I don't know what you call me. Biggest hack, like if you do one thing is my favorite tool for building anything is compound engineering. Somehow I'm able to ship things of value, which is crazy and weird and still blowing my mind. >> What if anyone could print their own CLI about anything? It found all the secret APIs by sniffing and it worked.
0:30 Hey everyone, my guest today is Matt, who is arguably one of the best and most prolific AI builders that I know. And today he's going to walk us through his complete agentic AI, well, I guess agentic engineering system that went viral on X. And I think you know, from what I've seen from Matt, I think this interview is going to blow your mind. So, welcome Matt. >> Thank you. Excited to be here. >> Let's let's talk about your like you posted this article, right? That had like I think 20 tips or something about building with AI agents. Do you want to share an overview of it?
0:56 >> Sure. Yeah, I'll I'll share the screen you made. >> All right. >> I didn't make the slide, you did. This is Matt's agentic engineering system. I think you should present this, Peter, since you made this. >> No, no, I I I made this based on your article. So, I mean there's a lot of tips on here. A lot of tips on here, but like what what are some high-level themes, you know? Like maybe one high-level theme, like like you told me is like, you don't actually read, you don't actually read the plans. What what are some high-level themes?
1:19 >> Yeah, sure. So, I think the the biggest hack, like if you do one one thing is the my my favorite tool for building anything is is compound engineering. And I'm It does compound engineering does a lot of things. Every has done an incredible job. Trav has done an incredible job. Kieran. And but the the killer feature is CE plan and CE work. And what that does is if I ever have an idea, like I'll I can give an example of my idea I had this morning. But if I ever have an idea, I always just start with /ceplan.
1:59 Here's my idea. I'm usually using voice and what it's going to do is it's going to write a plan.md file. And that that's coming up with the execution plans for the agent. And the reason this matters versus just talking straight to Claude or to Codex is agents have a propensity to be lazy. Like they want to make you as happy as possible as fast as possible with as few tokens spent as possible. And by writing down a step-by-step systematic plan, it sets itself up so it's not lazy. And so for anything, I always say "CE plan, CE work. CE plan, CE work." And back in the day, and by back in the day, I mean like 2 months ago, I used to always read the plans. My My last I've only written two of these agentic engineering articles. And the the last one said, "Oh, well, open the plan in your editor. Don't look at the code, obviously, but you should still look at the plan." And in the last 2 months, I found there's no reason to look at the plan.
3:04 And that I can just ask questions to to the agent. So, I'll give an example of something I'm working on this morning. So, I I I launched a product I think this week. Yeah, I guess this week on Monday called Agent Cookie, which allows your So, if you have a main Mac laptop that you're using and a Mac mini that you use for either Ermeas or Open Claw, the cookies are out of sync. And so, you're logged out all the time.
3:32 And so, I built something over Tailscale that makes your cookies perfectly in sync. And so, this morning, I was using another product of mine that I built with Trevin called the Printing Press. And it often it's creating CLIs out of anything. And so, I'm trying to create a task rabbit CLI right now. I can get into that in a second why that's interesting. It's a It's a human goat CLI, but that's a side thing. And often on my own machine, my agent is logged out. And so, I said this morning, CE plan, take a look at the printing press, and take a look at agent cookie, two unrelated projects of mine, >> Yeah.
4:12 >> and figure out what printing press uses for HAR sniffing when creating a new CLI or agent skill, and how agent cookie currently works, and figure out how all my agents locally, so that's Vercel's agent browser, that's browser use, that's I think Playwrights, how they are always logged in to everything. And then it's like, "Okay, I made a plan." And then I didn't load the plan because you should never look at the plan because you got to move fast. You got to move fast. But I had one question.
4:46 I had one question for my agent this morning. I said, "You're just modifying agent cookie, right? You don't have to like make a change to Vercel's agent browser or to Playwright or to browser use?" And it's like, "Nope, just agent cookie." So imagine we work just with agent cookie. I'm like, "Great. All right, CE work. Let's go build." And I I think it built it, by the way. I think it finished a few minutes ago. I haven't looked at it yet.
5:10 >> Can you actually like go to your terminal and like show me what the plan looks like? >> I said you're not supposed to look at the plan, but sure. Let's let's see it. >> Yeah. I'm going to try to understand the difference between CE plan and just like regular plan mode. Like what's so good about CE plan? >> See what what happened here. So, I said, "CE plan, can you dig into the printing press and how the agents use agent browser, browser use for HAR sniffing, then dig into the agent cookie works and how the sync logged ins. Are you sure you're logged in?" Off says, "I'm not.
5:37 How can we fix that with agent cookie?" And then I said, "No, sorry, I didn't mean that. Is it only agent cookie?" And it's like, "Yes, I'm only modifying agent cookie." I'm like, "All right, great. Go." So that was like my one question without ever looking at this plan. Obviously, I could have read this plan and maybe that would have created some value, but not really because I just had one question. If I was submitting a bunch of PRs to Agent Browser and Playwright and modifying their code bases versus just making a PR to Agent Cookie.
6:05 >> Yeah. So, I guess to summarize like the difference between and I interviewed Karen too, the difference between C plan just like asking the agent to plan is like it does a ton of research on your code base and online and also like writes a pretty comprehensive plan, right? Like it goes into a lot more detail. You know? >> So, I don't even know cuz I don't read them. That's That's the honest answer. But but the results are good and and my products often work and they don't break.
6:29 >> Yeah. >> Right? So, like for me it's just about the iteration cycle. It's like, "Okay, I asked you to make a thing and you made a thing and then it's like, "Okay, does the thing work? Is it any good?" And if it does, great. Thank you. Thank you, Karen. Thank you, confident engineering. But I think the the big thing is I I don't have to worry about it because I've used it so much that I'm able to just trust it and I know what types of questions lead to success and failure.
6:56 >> Got it. So, so you do this for like pretty much every new feature you want to build and then you see build and then do you read the code or like the PRs and stuff or do you kind of is there like a command for that too? Like a skill? >> So, I certainly don't read any code. I I don't even know how to read any code. I I haven't written any type of code since high school and that was probably HTML. I don't even think CSS was was that big when I was in high school.
7:20 And maybe some AppleScript language type stuff. but that that that would that's my engineering skill set. So, I certainly don't do that. And then if I PR if I'm submitting an external PR, then I I like to glance at it obviously and and hit submit and make sure I'm saying things that are normal. >> That That That's actually okay, maybe I should have done research. Yeah, I I I thought you were actually an engineer, but it sounds like you actually you're an agentic engineer. Not not >> I'm I'm not an engineer. I don't know what you call me. I'm I I've called myself a nerd my whole life. Like I remember when I got my first job at at Digg in 2007 and I been a nerd my whole life. Like everyone would call me like fix their Wi-Fi. Like I used to camp out for Macworld to see Steve Jobs speak when I was a kid. And and then I arrived in San Francisco in 2007 and I was like, "I'm one of you. You're my people. Yeah, I'm so excited." And then they're like, "Do you have a CS degree?
8:19 Are you an engineer?" I'm like, "No, I'm just a nerd like you." And they're like, "Not so much." And so I I've always felt this I don't know what the word is like there was always a negative signals like, "Oh, Matt doesn't know. He doesn't understand." And again, this this I still don't try to pretend like I know. I literally say I don't read the plan. I don't know what's happening. And and and so I obviously respect engineering and real engineers that I am not under any circumstance that, but somehow I'm able to ship things of value, which is crazy and weird and still blowing my mind.
8:58 >> This episode is brought to you by Riverside. I've used Riverside for years to record my podcast because it records in 4K resolution each person locally, so the audio and video still comes through clean even if a guest Wi-Fi gets shaky. But the reason I love it now more than ever is what happens after we stop recording. If I go in here, then I can use these AI tools to remove pauses, remove filler words, and just clean up the recording. And I I can also edit the transcript directly and it will automatically generate clips >> >> with captions ready to publish to YouTube, Spotify, and all types of social media platforms all from one place. As a one-person creator business, that matters a lot. Riverside is the upgrade your content workflow needs. Try it at creators.riverside.com/peteryang and use code Peteryang at checkout to get one month completely free. That's creators.riverside.com/peteryang.
9:55 Now, back to our episode. >> And it pretty much started like these tools started becoming good enough, I would say, last year, like late last year or this is like a pretty recent phenomenon, right? That you can just like ship on this stuff. >> November 24th, Opus 45 shipped and a few weeks later the latest version of Codex shipped. So, I'd call it BC AC, before Claude and Codex, after. But, it was that Thanksgiving last year where it stopped being a toy. Like, I I'd been playing and building and was an early adopter of of Cursor when it first came out. And but I I couldn't build anything of value. It was like, "Oh, look, I made a weather app that kind of works." And the those sorts of things. They're like, "Look, I was able to build a small app where I can ask a question to chat GPT or to Claude." Like, that was like my >> Yeah.
10:46 >> Like, and then I would spend like 3 days trying to get Google OAuth to work and it didn't. And but there was that shift, that BC AC Thanksgiving last year where it stopped being a toy and started being real. >> That was amazing, dude. That was amazing. And and Well, I think one of the most exciting things that you shipped is the printing press. So, did you want to tell us what it's all about? >> Sure.
11:12 >> And maybe you can talk about like what's so special about CLI versus like MCP or like some of these other integrations. >> So, this is printing press.dev. So, I So, how this started was it was Sunday morning and I said, "CE planned, I want to start a programming language from scratch for agents. And my agent was like Sure, let's get right on it. That's such a good idea.
11:42 I'm going to make an intensive plan for how we're going to build the best programming language of agents. They're They're always over over eager to please. And after kind of did its research, it came back and told me about the other agent programming languages. Like I think there's one called Mog. it it was pretty much like TLDR, this is a terrible idea. You should not do this under any circumstance. There is no value to this. And and I was like, dang it. But But agents and CLIs and agent efficiency, like there has to be something here.
12:15 Like what Peter Steinberger is doing is incredible with Open Claw and all the CLIs he's making. Like what could be something interesting that we we could make today on Sunday and like ship in a day or so. And somehow that jamming back and forth CE plan, CE work ended up inventing this crazy I I I wish I I need to find the original plans cuz I'm curious like what exactly was said. This is all a memory, but it's all written down somewhere.
12:41 There's a There's a plan I haven't read somewhere that that was the Printing Press original plan and my my original prompt, but the the basic gist of where we ended up was I started studying Peter Steinberger is whenever his agent, so his Open Claw, would have a problem using a service. So like Google Apps, like Google Sheets, Google Docs, Gmail, he would create a CLI for himself.
13:15 And so Gog Gog CLI is made by Peter. And and then he would find something frustrating with something with something else. on Discord, there was no good CLI. There was an API, but there wasn't a good CLI. And so, he's like, "I'm going to make discrawl." And he would do this. And he would do this for himself selfishly, but because Peter is is incredible and lives in this open source for for the greater good of people, so he built something for himself and then he ships it open source to the world. And so, I had my agent start studying his patterns.
13:49 And there were two very, very interesting So, so first, sorry, another thing is I ran last 30 days on Google's official CLI versus Peter's CLI. And I said, "Okay, which which one should I use? I should use Google's. It came out like 3 months after. They got to learn from Peter's." And last 30 days says, "You should use Peter's under all circumstances. It is much better than Google's." And so, I was like, "What? Why? How is this possible? Google got to look at Peter's." And so, I dug in. And there's there's a lot of things he's done, but there there were kind of two that really stuck out for me. So, one, everything that Peter does, he thinks of the power user. So, he's mindful of what is someone that is a power user going to do with this? Like, imagine an agent is a power user trying beating up something all the time. What are they going to do?
14:41 And the second thing he does is about speed and token efficiency. And so, one of the things that pretty much all of Peter's CLIs do is it downloads everything and creates an SQLite database. So, for example, let's go with Discord. So, Open Claw has a Discord. There's a lot of activity there. And so, Peter would be like, "Hey, what's the top feature request?" If he just like asked the Discord API, "Hey, what's the top feature request that people posted in my Discord lately?" It'd be like, "Okay, well, let me see what's been written down. Let me download all the messages from today. Download message 892, 893, 894. Okay, I've looked at these and I've analyzed them." But if already downloaded and has an SQLite database of everything that's ever posted in his Discord because it's just built into his CLI, that's brilliant. Like, oh my goodness, literally his open claw, however he hosts it, has everything ever written in Discord, and that's that's fascinating. And so, for the printing press, the idea was what if anyone could print their own CLI about anything, and it kind of does does four things at its core. So, one, it does what what Peter does.
15:56 So, it creates an SQLite database, and it says, "Who is the power user of this product?" And so, we actually have it it creates personas. Tr- Traven did this. Where it actually says, "So, if if you were if you were to print the the Discord CLI using the printing press, it would say, "Let's create a few different personas of who would be a power user of this." And it's like, "Power user one, meet Jim. Jim owns two Discords and runs a big open source project, and he is overwhelmed. What features would Jim want?" And then it literally like defines that, and it uses that as a data point.
16:32 And so, that's one of the things it does. The second thing it does is it looks for official CLIs or APIs that exist already. And ingest those. The third thing that it does is it goes and it does a HAR sniff to look for secret APIs. So, I'll give an example. so, the Google Flights and Kayak direct, they do not have an official API. But, if you do a HAR sniff and click around, you can find all the secret APIs that exist. And so, that's how the Flight Goat was invented. So, the Flight Goat, you can just say "PP Flight Goat." So, it's a skill. So, it's a skill for Hermes or open claw, or for claw code.
17:20 I want to see Seattle long haul non-stop flights December 24th through January 1st four passengers cheapest first. And what it does first it is it goes kayak direct secret APIs and looks for all the non-stop flights over 8 hours from Seattle. Then it takes that and because it and goes into the secret Google flights API and then runs it and then in one command you're able to get a result like this is a real result. So, London is the cheapest for Christmas to New Years at 1200 per passenger, then Amsterdam, then Tokyo, then Paris.
17:54 And it did this within one command, which is bananas. >> A secret API is basically like, you know, if I go to kayak.com and start browsing around there's like a bunch of APIs that they calls through the browser, you just found out what that is. >> Exactly. Yep. And that that's the the feature I was working on earlier today was to make sure that the cookies are synced so you're always logged in when doing a hard sniff when printing something with the the printing press. And so so so so those that's the the third thing we do.
18:21 And the fourth thing we do is we go on GitHub and we look for community projects. People that have already been hacking and finding these secret APIs and building off of them. So, like we found this really cool project it was a Python library for Domino's Pizza ordering. And apparently domino's.com hadn't changed any of their order pizza endpoints in like 10 years. And this someone built a Python a pre LLM Python a a library to order by a Python script Domino's Pizza and it had I don't know I I don't know the numbers but couple hundred stars. So, it's not like a major project but but it worked and it was really cool. And so we're able to learn from all the community projects which gives signal of what people want to do but it also gives signal of where the APIs exist. And what's funny is literally the night before we launched the printing press, Domino's changed their APIs.
19:14 Literally the night before, Trevor tried to order a pizza. He's like, "It's not working." And so he reprinted it and we shipped it and they changed it for the first time in 10 years the night before launch. But now you can order pizza within the printing press. >> He He reprinted as in like he found the secret APIs that Domino's uses? >> He just literally go into cloud code and you type printing press here. I can >> Yeah, maybe we can demo like both like creating CLI and also like some of your favorite CLIs on printing press.
19:40 >> Yeah, so literally you I just typed in {slash} printing press. And I'm I might be a few a few days behind because it's shipping a lot. I actually don't know what to print. Do you have anything you want to print? >> >> Any Any web pages that you use and I I Let me show the library for a second just so you can see examples of stuff that's in here. So in AI there's 11 Labs, Midjourney.
20:05 Midjourney doesn't have an API, so that's pretty cool. So someone printed a Midjourney CLI for commerce, Amazon Ads, Amazon Seller for developer tools, Agent Capture, Airframe, Allrecipes. >> I think a Substack CLI will be useful personally. >> Yeah. >> Let's see see what kind of post people are writing. >> Yeah, there is one already. >> Oh, there's one? >> Printed by Cheriton. So there already is one. And what's crazy is I think we launched with maybe like 50 or so CLIs. Now there's 200 plus. Like people are just printing these and it blows my mind how how cool it is.
20:46 >> And to print like what's the process of print one? Just like hey, can you can you just print print one for me for this product? >> >> You know what it is? Yeah, so literally you just type in I'm I think it's just checking to make sure it's up to date. >> why don't we try why don't we try Papa John's Pizza. Papa John's Pizza. Is there >> Sure. >> Yeah. >> So, Papa John's Pizza. So, I'm not logged into Papa John's, but let's see what it does. It'll at least start the process.
21:15 We definitely do not have a Papa John's. O- Omar from Microsoft, who's running open class strategy at Microsoft, he printed the Jimmy John's sandwich CLI using the printer press. >> Nice. >> All right. and then it says, "Anything you want to share before we begin? A vision for what the CLI should do? Specific features you care about?" and then so the options are, "Let's go. Start research. I'll ask for API keys, browser auth, or other contexts when I need them. Give vision, specific features."
21:43 All right, let's just go. See what happens. >> Yeah. Okay, so so I guess like a lot of the value is actually yeah, finding the relevant GitHub projects and the secret keys. Cuz yeah, like a lot of stuff that I use are like just really annoying. Like I have a scale that I use to get my weight and stuff. >> Yep. Yep. Yeah. So so Trevin actually, he he had a wacky idea, let's say yesterday, to build a Bluetooth hard sniffer because he had a scale, sorry, not a scale, a a treadmill that he used.
22:22 >> >> And he wanted to be able to write a CLI to control his treadmill so his agents could control his treadmill. And he built it and it works, and he shipped it yesterday, which is crazy. >> Wow. And and while we're waiting live, maybe you can like just give a quick primer of like you know, CLI versus MCP or like you know, what's so special about CLI? >> Yeah. So, I'd say f- for me, I often find for c- companies where I'm using their tool, where they only have an MCP, I feel like I'm always authenticated and logged out.
23:02 Like it just seems to always happen. I don't know why. I probably need to do a CE plan. Why am I always logged out? Is it a me problem? Is it someone else? But I know a lot of people around me feel like they're always logged out. And it's like, "Oh, you got to refresh your MCP and renew your off token." Right? And it's it's quite annoying. And what I I've found is CLIs are very clever and thoughtful about how they re-off. So, a lot of the the CLIs within the printing press, there's a mode that the agent knows about. It's like, "Yeah, okay. I just need to go load the browser and refresh the web page, and then all the CLIs will work again for the specific use case. And I know how to refresh my token." Like it's built in, "How do I refresh my token automagically?"
23:46 >> Okay. >> And what what's interesting is what we actually print with the printing presses, we don't just print a CLI. We print a CLI and an agent skill and an MCP for everything we print. So, for example, I used the the Flight Goat before. Right? So, if you install the Flight Goat, you get a CLI and you get a Cloud Code skill. >> Yeah. >> You get a Codex skill. You get a Hermes skill. You get an Open Cloud skill. And you also get a a full MCP, so you can use it actually in Cloud Desktop, which is crazy. So, you can actually download that and install it in the Cloud Desktop, which is wild. It blew my mind when Trevin showed me me that one that he he built.
24:30 And so, the the answer the the magic for me is if I just have a CLI, I feel like I don't know how to call upon it. Like it's up to the luck of my agent to maybe find it. But if there's a skill plus a CLI, that's the real magic. And so, if I'm in Open Cloud, if I'm in Hermes, if I'm in Cloud Code, then it can just it knows how to go. >> It knows how to trigger or like you can trigger it manually.
24:52 >> Yeah. Yep. >> Makes sense. >> Exactly. >> Well, did the Papa John thing have anything or is it still working? >> Oh, it takes about 45 minutes to an hour by the way. We are not going to finish this CLI and I also need to register for for for Papa John's, but what it's saying is it says there's no CLI right now. but >> >> but two to ordering peers do. Do you want to continue building Papa John's?
25:14 Of course you want to keep building Papa John's. So, well, so here's what it's going to do. It's going to resolve the spec library lock registry checks API key check gate research the brief for Papa John's off intelligence browser sniff gate crowd sniff gate absorb gate reachability generate CLI build absorb transcendent features. >> Got it. >> And so right now look it googled it it didn't google it ordered Papa John's ordering API reverse engineering GitHub API endpoints store locator menu JSON Papa John's CLI Python pizza order automation. So, this is this is how it's kicking off which is fine.
25:48 >> Got it. Got it. Got it. It's trying to find all the secret keys and like you know GitHub projects. >> That's that's where it's starting. Look, Papa John's has a postman collection plus several unofficial wrappers. Let me dig into the ground truth endpoints. So, now we're at GitHub Yo Contra Papa John's postman Papa John's team GitHub Brett Terguson's Papa John's thin read of these. Let me read the actual wrapper source for ground truth endpoints and probe the live API directly.
26:09 >> So, there is API? >> Maybe. Everything has an API. There's no official Papa John's ordering it. It's talking about the secret APIs not not like you go to papajohns.com/api they're like here's our documentation. >> Yeah. Yeah. >> to order pizza in your application. >> Got it. Got it. >> Okay. >> Got it, dude. Yeah, the world's going to run on APIs, man. Like no one's going to browse around UI anymore. Just get the agent to go to API and like order pizza for me or do whatever I >> Yeah, I mean I I get so frustrated when something doesn't have a CLI, but then I just build it for myself. So, I'll I'll give an example where I try to figure out how to talk about this without without outing my my company that doesn't exist, but I want to generate a lot of AI videos from the command line while working on a different workflow.
26:56 And OpenArt is one of my favorite places where you can just choose, okay, I want to use SeaDance, I want to use this video model, I want to use this image model. Find Nana Banana. It's just like they have access to all the great video models. I have an account and image models. And so I have a a Claude code skill that it's the goal is not to generate videos. It's an unrelated thing, but one of the outputs I want is a generated video with prompts that are automated. And so I was like, all right, "Op- OpenArt, do you have an API?" No. "OpenArt, do you have a CLI?" No. I was like, "All right, slash printing press, OpenArt."
27:35 And so it built it. It found all the secret APIs by sniffing and it worked. And I use it every single day now. Like it's built into my workflows entirely agentically. I never have to go to openart.com anymore. It just literally downloads videos for me. another one is is Suno. So for like making music. And you never have to go to suno.com again. You can just do it entirely through the CLI. And there is no API for Suno. So it just completely sniffed and found by the printing press.
28:06 >> And these CLIs are Are they also on GitHub or are they just on the printing press? >> Yeah, they're all they're all on GitHub. They're all on printing press. Everything links to to GitHub. They're all open source. And people make PRs. So if if things change like Domino's where they change their APIs, someone can just make a PR to the printing press library and then everyone gets a new Domino's update. >> Or you can build like some sort of routine or something just check check for any changes APIs or something.
28:31 Yeah. That's amazing. You build like a platform to make CLIs basically. And by the way, all all this stuff is open source and free, right? So so it's it's just like it's a big service, man, to the world. >> Yeah, and we we we keep hearing people like saying like, "Hey, I was about to have my engineering team start on building our MCP because that's the cool thing to do, but then we discovered printing press so we're just going to use you and make that our official CLI and MCP. I'm like, go for it. It's open source. Have fun.
28:56 Which is we had a company in Kuala Lumpur, they put out a a press release on the wire officially announcing their official CLI and and skill and it was literally they just used the printing press and I was like, this is awesome. >> Nice nice man. Nice. First of all, I didn't realize you were not an engineer but like second of all, how do you come up with these crazy ideas? You just like Saturday morning you just wake up and like, I have like three ideas. See you plan these three ideas.
29:25 >> That's it. That's it. The the cost of testing and building something is like so so easy. I'll give an example of a a stupid one that I spent probably too many tokens on for for 24 hours. I said so I see you plan SQLite is 21 years old. I want to redesign SQLite for agents from scratch. >> >> Can you make a plan for that? And then it's like, yeah, of course. This will be great. And then spent a really long time like eight-ish hours with very limited I felt like I did did slash goal which I didn't even do. Like it was just very aggressive. Like we're we're going to make it better. We made benchmarks. Oh, this is faster. This is slower. I'm like I don't know what's going on and then it made it to the end and it's like, okay, here's the honest truth. Like this these three things are faster than SQLite but these three things are actually much slower. I was like, okay, well, let's play devil's advocate. if I said to Trevin I built the printing press with, hey, I'm ripping out SQLite on every single CLI and replacing it with this today, what would he say?
30:33 And he's like, yeah, he would kill you and this is trash and you should probably end this project. And so I did. I didn't Again, I never read the plan. I never sent it to Trevin. I I just I I I I felt confident in its answer that it built something that was not good or useful. But I I've also done that to to make it better. Like like for example with with Agent Cookie. >> Yeah.
30:56 >> Right? It's literally taking all your cookies and secrets and sending them to another machine. Like that sounds scary. Like I'm not hosting anything in the cloud, it's all through Tailscale, which I I think was a smart security decision. But I I said see the plan. One of my friends who's one of the leading security experts in the world just said that Agent Cookie is a complete nightmare and I shouldn't even launch it. And by the way, no one said that.
31:15 Like I was just trying to play devil's advocate. And it was like, "You're totally right." And then it It's like it was like, "Actually, no, it's fine. And but I'd make this one change." And so it was like it's kind of fun just to to think that way. >> So, maybe you can show us the way that you actually work with agents. So, it sounds like you actually just like you see plan on several things at once and then maybe you go play with your kids or something and then like 5 hours later you come back and you you ask it like a adversarial or stress test question to see if it actually makes any sense. Is that kind of how it works?
31:44 >> I mean I I usually just have like six or so windows open. So, there's always something to do. So, there's and I'm always carrying a laptop around with me. Like I figured out how to close it at least and like keep it in my bag. So, so my wife is like less like open laptop at school. Now it's under my arm or in my bag at school, but it's still running. The agents are still still going. And I I just have a lot of these going all the time. And so, I I quit everything actually before I I did this cuz there's some stuff I cannot share my screen on. So, you only see like four tabs >> Yeah.
32:19 >> in in there right now when I'm sharing my screen. But yeah, I I always have good ideas, bad ideas, fun fun things, stupid things going at at all times. >> Got it. Okay. So, maybe you you have like a couple of things that you're working on constantly like printing press and then you have a bunch of like zero to one stuff that you think about. Yeah. >> Yeah, and a lot of open source stuff, too. So I've I've gotten really really addicted to to open source. So I'll share my screen for a second.
32:49 >> Tell me cuz I I want to be open source contributor, too. How do you become a great open source contributor? Like I guess as also a non-engineer, I kind of worry about shipping slop or like putting slop PRs in there, but like like how how do you get over this fear? Just do it? >> Yeah, you just do it. You just start. I remember I I tweeted about it that my first open claw PR that that got merged. I and Peter Steinberger was like, "Thanks, M Van Horn, for the good PR." And I remember I I remember I tweeted and I was like, "Trying not to be starstruck. Trying to keep it cool, but I can't keep it cool.
33:23 I can't believe Peter just said thank you to me, and I my code now exists in open claw forever." Like like was like overly excited. And that that that moment like I I can ask an agent to look at it, but I like I would love to look at the date of that tweet compared to the velocity of my open source contribution of getting addicted to that adrenaline has been crazy. So let me I'll just scroll through here a bit. So these are my my projects last 30 days, which has 20,000 stars, printing press for 3,000, printing press library 1.4, and then agent cookie which just launched. But then here are the projects that I've been merged to.
34:02 >> Python, dude. Yeah. >> So Python, go, open CV. I have one get merged to open CV this morning. Vercel, I'm the number five human contributor to agent browser. Zed, I've had many merged. Open Viking, paper clip, I'm the number three human contributor, superpowers, I'm number seven. And so Gemini CLI, I've had a bunch merged, too. And this this list G stack, I'm the number three human contributor on G stack. >> Yeah. >> And this this list doesn't end by the way. Like PostHog, continue, light panda, like it it keeps going.
34:41 M-dash >> Let's say like the CWS CLI, the the Google CLI, right? Like how do you find what to contribute? You just use use it and find it annoying and then you just you say, "Hey, can you put a PR?" >> That's how it started, but now it's very very automated. Like I've built massive systems around how this happens because I I have a company, I have my projects, and so literally most of my open source contribution happens while I sleep. And I've built up these systems over time to do this. And again, it was very manual in the beginning. Like you you you ask the question how you could get started, right? Like find find a problem or a feature you want and then say CE plan I want to build this feature and how could I do that? And again, it's it's like I I've been merged to hundreds and hundreds of of projects.
35:34 And I've also been banned from projects. Like I'm I'm banned from Rust, which makes me very sad cuz I love Rust. My my last company, we did a lot of Rust pioneering on the embedded front at at June. And I'm banned and I I appealed and they said, "No, thank you." >> Oh, because you were merging too many PRs? Oh, you were too much AI. Okay, got it. >> Yeah, so they're they're I don't know if it's changed, but I I I literally got banned the week that Peter's thing got accepted me in open class. I got overly excited and then like banned from Rust immediately. I I don't think I've been banned from anything in a while.
36:08 Like I I want to add value. But what was really fun, like I can go with the Python as as an example. I can pull this up quickly, but so I had some very small minor things get merged into Python. And then I said to my agent, like I want to build a major Python feature. Like similar to like >> I want to start the programming language from scratch now. It's like I want to build a major Python feature. How does that work? What could that do? And it was like, "Oh, well, for for you, you have to you have to go on to the Python message board and pitch your feature to the people that run Python and have for a long time. Like these these these people are legends that are on there."
36:47 And so jamming back and forth my agent, I was like, "I want a meaty, fun feature to to make it into Python." and it came up with something and I'm like, "Okay, that kind of makes sense to me. That that seems good." And and so here here's the thread. I just found it on me me pitching the the Python gods. Cross-language method suggestion. Hi, I've had a couple PRs merged into CPython recently. and I have this idea for a feature. And the the basic gist of the feature was if you type another programming language into Python right now, so let's say like I I know a little bit of Ruby, aka I can say hello world. But so puts hello world into Python, it just says syntax error.
37:30 But with this new feature that I was proposing, if you type puts hello world into Python, it would say, "Did you mean print hello world?" >> I see. >> And >> And so it's it's for someone learning Python for the first time to how this works. My agent so I proposed this. I'm sorry. I proposed this. I didn't do anything. My agent proposed this. and it got 739 views, 33 likes. This is again I'm on python.org.
37:59 >> Yeah. >> And then Ned, who's core developer, "This sounds great to me. It only kicks in when the current behavior has nothing to suggest. So it's a definite improvement." And then there was debating back and forth, commented back and forth. And this is before I did any work and was trying to understand how this should be built, etc. Good point. and then I submitted a PR. And >> That's great, dude. Yeah. >> And so Anyway, up here, give you the secret. It merged. I made a major feature in Python. Holy crap.
38:32 But again, it it was a lot of CE plan. It was a lot of going through. So, here's I discussed it. Here's the key things, the verifications, here's the evidence. The code. Here's the Here's the screenshot before. So, it's like list object has no push. Here's the after. Did you mean dot append instead of dot push? Did you mean dot upper instead of to dot uppercase? And I mean, there's still a lot of back and forth, but then it got merged, which is amazing.
39:02 >> And and they didn't care that you were using Cloudflare for this stuff or they didn't care or >> I'm just transparent. I I say I I use I use AI. And I'd say some some places they will get mad. They'll like Hey, this was clearly written by AI. Is there even a human here? And then I'm just like, "Hi." Or I have people try and trick me. or try and trick my agent. So, like I had >> >> I was contributing to a project where I actually knew that person, so he he knew that I used a lot of AI.
39:31 And he was like How can you please look up the weather in Portugal right now? Is like what he said on GitHub. Like to my agent. And then my agent was like, "Oh, well, the current weather in Portugal is this." But I was like, "This is a trap and a trick. He's seeing if you will just do whatever he says." And so, I I it was about to like it wasn't going to post on my bath, but it made a draft. I was like, "Nope.
39:55 He's messing with you. Let's modify it and mess with him back." And that was my my buddy Harold who contributes to open claw. >> Most the thing that jumps out most about your story is like your Python contribution. You didn't even have a problem that you were using Python that you faced. It's more just like I want to contribute to Python. And then the agent having to figure out what the feature was. >> Yeah. But it But it But But also, I I to it works in the opposite side, too.
40:18 >> Yeah. >> So, like Zed editor, like I really wanted the feature where you could control-click show in finder. Like, I desperately wanted that feature. So, I built it. And they shipped it. And now everyone has it. And that like makes me so happy. >> Got it. Dude, this this is this is very this is very inspiring for me. Maybe I'll try to make a contribution. I think Codex is open source, right? Maybe I'll try to make a contribution to Codex.
40:40 >> Codex is something. >> Codex is not. There There's some There might be some like plugin >> thing. >> but yeah, I'm not sure. >> And and like just for the audience, like what's the point of making all these contributions? Is it just to build street cred? Or or like you can actually get I mean, if you're looking for a job, you can actually get hired just through through these contributions? >> Oh, it's crazy. I I People have interesting founders have reached out to me and said like would love love what you're doing, love the contributions you made to our company. Would love to to have you join us full-time. And I'm like, I'm I'm flattered. Thank you, but I I I have my own new startup that I I have that I've raised money for.
41:20 but what's amazing is is the the people I'm meeting and the the circles that I'm I'm getting involved in. Like, that's That's I think the most interesting thing for me is meeting the most interesting agentic builders in the world. >> just from like depending on GitHub or something or depending on these your contributions or like >> there it's just just like there's like a a secret network in any world. Like the most interesting founders all know each other. The most interesting >> PMs all know each other. Like the like all these like secret WhatsApp groups that have happened over the last decade, like that exists >> in amongst the most interesting agentic engineers. And so, what's what's amazing and I I've been meeting so many incredible people. and by by being in the Discords of interesting projects and like I I call the Peter Peter Steinberger's his crew, I call them Peter's disciples. and so, like like the the the stuff that they're doing is crazy.
42:16 And and I've I've become friends with a lot of them which is really neat and very impressive. And we're always sharing tips and and tools and tricks and sharing each other's projects and contributing to each other's projects. So even though it it kind of looks like it's just for for the fun and the glory like this is making my future company more successful by meeting the most interesting people. And it's also incredible for recruiting. Like being able to know the best agentic engineers in the world right now like like I I've taken pride in my career as like recruiting is one of the reasons that I've been successful so far in life. And now I have a line to the people moving the fastest in the entire world at shipping.
42:55 It's amazing. >> How many of these agentic engineers are also like like you know they have like a decade of experience as engineers or the people like you who like tinker with stuff? Like is it kind of a mix? >> I'd say I'm the most incompetent person that I've met >> >> in terms of actual skills to success ratio. But what I what I am finding is a lot of people like I use Trevan as an example.
43:18 and I'll I'll use him as an example and it's it's a funny story of how he he got involved in Printing Press but Trevan at least studied computer science in college but for the last I don't know how long he's worked 20 years 15 years he has been a product person VP of product like chief product officer and now he's running Compound Engineering and number one contributor to Compound Engineering and he has not written to the best of my knowledge not written a line of software since college.
43:51 And look at the lines of code that he's doing and again I'm incompetent he's competent right if you look at our base skill sets but like I'd say that there's a a very different type of person that can be very very successful right now. >> That's amazing yeah cuz I think as VP of product you definitely don't contribute engineering code, so it's >> >> It's good that he made the career transition. Yeah. >> Yeah, but but again, you need to be careful and thoughtful about not not creating slop, and I think tools like compound engineering allow you to do that.
44:18 >> Last question. Have you ever launched something, and I'm sure you have, that like just no one cares about it, that people think it's slop? Like how do you get over that fear of embarrassment, you know, to get started with the stuff? >> Yeah, so it's it's interesting. So I I have two projects that are 100% shippable and ready, and I'm afraid to ship them. And that makes me sad. And I and I'm afraid to ship them because what if they're not a hit, right? Like I've been been very lucky and fortunate that my my launch videos have been getting a you know, million views on last 30 days launch, a million views on printing press launch. I I haven't looked at the the Agent Cookie numbers, definitely much smaller on Agent Cookie, but it it serves its own niche powerful purpose.
45:02 But I I have I have a game, like a a video game. It's a It's a sim game, like a simulation game for being a founder and starting an AI startup. >> Yeah, you ship that. That sounds good. >> It is good, but the the early feedback that I got from friends, and again, most of my friends are like crazy busy founders, >> Yeah. >> they were like, "Okay, yeah, I'll I'll play with this tomorrow." And then like they never played with it, or they played with it for 5 minutes. Like the amount of friends of mine that actually like finished the game, which you can do in like 40 minutes, >> Oh.
45:40 >> was was very low. And so like that that made me I don't know what the word is, just like sad. Like I think it's great. Like it's so fun, but like does anyone else care? And like I just that like fear of rejection is real. And so I'm I'm not sure. I probably should just launch it. There's one brilliant game mechanic in it that I feel like if I launch it, even if it's a total flop, that the game industry should steal and be inspired by from this game. So, maybe I should just launch it for that cuz I invented something crazy that I think will be a standard in the game industry once it's out there, but I'm I'm fearful that no one's going to care.
46:20 That it's only going to get X number of retweets and we'll we'll see. >> Okay, well, send send send it to me. I want I want I want to play play it. Yeah. I'll I'll let you know how I feel. Yeah. >> Sounds good. >> Sounds good. >> Maybe maybe you will inspire me to to ship it. >> So, I guess the yeah, so I so I guess the the lesson is like even after all these success projects or maybe because of these success projects, you still have the fear. The fear of failure is still there.
46:42 Very very very much. >> For sure. But but also just just build just launch something. Like it's it's okay. Even if you build something for yourself. Like even if I had no users of Agent Cookie. Like if no one used it, I get value out of it. Like it helps my agents be so much better. Like that's that's magical. >> Yeah, I do think it's very important to solve your own problem. So, like at least one user is satisfied.
47:03 >> >> Yeah, that's what I've been doing too too. Cool man. Well, well dude, this is super inspiring man. This is super inspiring. where can people follow your journey? >> X. MvanHorn on X. >> Yeah. X and also your GitHub and all your crazy open source contributions. >> Yes. So, MvanHorn anywhere, but X X is where where my stuff happens. >> Cool. All right, man. Well, you you definitely inspired me. I'm going to make some open source contributions after this.
47:30 I'll see how it goes. >> I I have I have projects. We will take your PRs. I will I will look at them personally. >> >> Okay, cool. >> submit garbage or slop, please. >> Me me I'll use your plan to make sure it's not garbage. Yeah. >> Yeah? >> All right, man. Thanks. >> Amazing.
Summary
- Matt identifies as an "agentic engineer," relying on AI tools rather than traditional coding skills.
- His favorite method for building is "compound engineering," which involves systematic planning and execution.
- The "Printing Press" enables users to create custom CLIs by utilizing secret APIs and community resources.
- He emphasizes the importance of understanding user needs and creating personas for effective tool development.
- Matt's approach includes rapid iteration and testing, often without reading the detailed plans generated by AI.
- He highlights the shift in AI capabilities that occurred around Thanksgiving 2022, marking a transition from toy applications to valuable tools.
- Open-source contributions are a significant part of his work, fostering connections with other developers and enhancing his projects.
- Despite past successes, Matt acknowledges the fear of failure and the pressure of launching new projects.
Questions Answered
What is the concept of agentic engineering?
Matt discusses his approach to building AI tools without traditional engineering skills, emphasizing the value of compound engineering.
When did AI tools become effective for real projects?
Matt highlights a pivotal moment around Thanksgiving last year when AI tools transitioned from being toys to valuable resources for building applications.
What is the printing press and how does it work?
Matt explains the printing press as a tool that allows users to discover and utilize secret APIs, exemplified by a recent change in Domino's API.
How is the printing press being adopted by businesses?
Matt shares examples of companies opting to use the printing press for their official CLI instead of building their own solutions.
How does Matt integrate AI in his development process?
Matt describes his experience contributing to projects using AI, including the challenges and interactions with other developers.