You Can't Run AI Agents Without This

The fastest way to make an AI agent The fastest way to make an AI agent dangerous, I'm convinced of this, is to dangerous, I'm convinced of this, is to dangerous, I'm convinced of this, is to let everyone use it and nobody own it. let everyone use it and nobody own it. let everyone use it and nobody own it. And I want to start there because I And I want to start there because I And I want to start there because I think we've made agents sound way more think we've made agents sound way more think we've made agents sound way more confusing than they really need to be. A confusing than they really need to be. A confusing than they really need to be. A lot of us hear the word agent and we lot of us hear the word agent and we lot of us hear the word agent and we immediately think, okay, this is some immediately think, okay, this is some immediately think, okay, this is some fully autonomous thing running in the fully autonomous thing running in the fully autonomous thing running in the background. Is this a digital employee? background. Is this a digital employee? background. Is this a digital employee? Is this a model? Is this codeex? Is this Is this a model? Is this codeex? Is this Is this a model? Is this codeex? Is this claude? Is this chat GPT with a custom claude? Is this chat GPT with a custom claude? Is this chat GPT with a custom trench coat? What are we talking about? trench coat? What are we talking about? trench coat? What are we talking about? And that confusion matters because the And that confusion matters because the And that confusion matters because the word agent means kind of all of those word agent means kind of all of those word agent means kind of all of those things in different contexts. And if things in different contexts. And if things in different contexts. And if you're still stuck asking, "Am I even you're still stuck asking, "Am I even you're still stuck asking, "Am I even using an agent?" You are probably not using an agent?" You are probably not using an agent?" You are probably not asking the most important question. asking the most important question. asking the most important question. Who's responsible for the work that this Who's responsible for the work that this Who's responsible for the work that this thing is now doing? So, this video is thing is now doing? So, this video is thing is now doing? So, this video is going to be very simple. I want to make going to be very simple. I want to make going to be very simple. I want to make the basics clear. How to know if the basics clear. How to know if the basics clear. How to know if something is close enough to an agent something is close enough to an agent something is close enough to an agent that you should care. What to do with it that you should care. What to do with it that you should care. What to do with it once you have it, what care and feeding once you have it, what care and feeding once you have it, what care and feeding means, and when this becomes a team means, and when this becomes a team means, and when this becomes a team workflow. Because the big shift here is workflow. Because the big shift here is workflow. Because the big shift here is not that everyone needs to become an AI not that everyone needs to become an AI not that everyone needs to become an AI engineer. The big shift is that more of engineer. The big shift is that more of engineer. The big shift is that more of us are going to have little systems that us are going to have little systems that us are going to have little systems that do work for us. And if those systems can do work for us. And if those systems can do work for us. And if those systems can read files and draft messages and maybe read files and draft messages and maybe read files and draft messages and maybe change code or summarize customers or change code or summarize customers or change code or summarize customers or update records, somebody needs to own update records, somebody needs to own update records, somebody needs to own them. Not philosophically, not on an org them. Not philosophically, not on an org them. Not philosophically, not on an org chart operationally. So let's make this chart operationally. So let's make this chart operationally. So let's make this really concrete. Let's say you open Chad really concrete. Let's say you open Chad really concrete. Let's say you open Chad CPT. You can go ahead and do that. Type CPT. You can go ahead and do that. Type CPT. You can go ahead and do that. Type a question, get an answer, and then you a question, get an answer, and then you a question, get an answer, and then you move on about your day. That is an move on about your day. That is an move on about your day. That is an assistant interaction. You asked, it assistant interaction. You asked, it assistant interaction. You asked, it answered. You decide what to do next.

answered. You decide what to do next. answered. You decide what to do next. Now, let's say you have a custom GPT Now, let's say you have a custom GPT Now, let's say you have a custom GPT that reads your notes every week. It that reads your notes every week. It that reads your notes every week. It prepares your Monday priorities. It prepares your Monday priorities. It prepares your Monday priorities. It follows your rules. And it produces a follows your rules. And it produces a follows your rules. And it produces a work product you actually use. That is work product you actually use. That is work product you actually use. That is close enough to an agent for this close enough to an agent for this close enough to an agent for this conversation. Let's say you open Claude conversation. Let's say you open Claude conversation. Let's say you open Claude and ask it to help write a paragraph. and ask it to help write a paragraph. and ask it to help write a paragraph. Same thing. That's mostly an assistant. Same thing. That's mostly an assistant. Same thing. That's mostly an assistant. But if you have a clawed project with But if you have a clawed project with But if you have a clawed project with files and instructions and examples in a files and instructions and examples in a files and instructions and examples in a repeated job or if you use claude code repeated job or if you use claude code repeated job or if you use claude code to inspect files and make changes and to inspect files and make changes and to inspect files and make changes and issue commands and come back with a issue commands and come back with a issue commands and come back with a result, now you're in Asian territory. result, now you're in Asian territory. result, now you're in Asian territory. And many uses of codecs are very clearly And many uses of codecs are very clearly And many uses of codecs are very clearly in this category. If you give codeex a in this category. If you give codeex a in this category. If you give codeex a repo and say inspect this code, fix the repo and say inspect this code, fix the repo and say inspect this code, fix the bug, run the test, show me the diff, bug, run the test, show me the diff, bug, run the test, show me the diff, it's doing work across steps with tools it's doing work across steps with tools it's doing work across steps with tools and real consequences. It may be and real consequences. It may be and real consequences. It may be supervised, it may ask for approval, it supervised, it may ask for approval, it supervised, it may ask for approval, it may not be autonomous in the sci-fi may not be autonomous in the sci-fi may not be autonomous in the sci-fi sense, but it's an agentic workflow. The sense, but it's an agentic workflow. The sense, but it's an agentic workflow. The brand name, the word agent is not the brand name, the word agent is not the brand name, the word agent is not the point. Chad GPT, Claude, Codeex, Cursor, point. Chad GPT, Claude, Codeex, Cursor, point. Chad GPT, Claude, Codeex, Cursor, a workflow tool, they can all be agents. a workflow tool, they can all be agents. a workflow tool, they can all be agents. The label doesn't matter. The job does. The label doesn't matter. The job does. The label doesn't matter. The job does. And once you delegate a job, your job of And once you delegate a job, your job of And once you delegate a job, your job of owning starts. And I think this is where owning starts. And I think this is where owning starts. And I think this is where people get tripped up because the AI people get tripped up because the AI people get tripped up because the AI conversation is still obsessed with conversation is still obsessed with conversation is still obsessed with building. Build an agent, make an agent, building. Build an agent, make an agent, building. Build an agent, make an agent, launch an agent, connect tools, automate launch an agent, connect tools, automate launch an agent, connect tools, automate the workflow. And sure, that matters. I the workflow. And sure, that matters. I the workflow. And sure, that matters. I like building. I want people to build.

like building. I want people to build. like building. I want people to build. But the moment after the initial build But the moment after the initial build But the moment after the initial build demo is where your real responsibility demo is where your real responsibility demo is where your real responsibility for that agent over the long term for that agent over the long term for that agent over the long term begins. Useful agents don't stay in demo begins. Useful agents don't stay in demo begins. Useful agents don't stay in demo land on day one and two. They start land on day one and two. They start land on day one and two. They start becoming part of how daily work gets becoming part of how daily work gets becoming part of how daily work gets done for you for your team. A research done for you for your team. A research done for you for your team. A research agent has to find sources you trust agent has to find sources you trust agent has to find sources you trust every single day. A writing agent has to every single day. A writing agent has to every single day. A writing agent has to work with your voice as it evolves every work with your voice as it evolves every work with your voice as it evolves every single day. A coding agent changes files single day. A coding agent changes files single day. A coding agent changes files and has to be trusted to do that. A and has to be trusted to do that. A and has to be trusted to do that. A support agent shapes what customers hear support agent shapes what customers hear support agent shapes what customers hear from the company. That's critical. A from the company. That's critical. A from the company. That's critical. A product agent shapes what shows up in product agent shapes what shows up in product agent shapes what shows up in the backlog for engineering. And if no the backlog for engineering. And if no the backlog for engineering. And if no one owns that agent and feels like they one owns that agent and feels like they one owns that agent and feels like they have skin in the game, the danger can have skin in the game, the danger can have skin in the game, the danger can feel ordinary until it's not. The agent feel ordinary until it's not. The agent feel ordinary until it's not. The agent will use an old policy. Maybe it'll pull will use an old policy. Maybe it'll pull will use an old policy. Maybe it'll pull from stale docs. Maybe it'll repeat a from stale docs. Maybe it'll repeat a from stale docs. Maybe it'll repeat a bad pattern. Maybe it will draft bad pattern. Maybe it will draft bad pattern. Maybe it will draft something plausible but wrong and you'll something plausible but wrong and you'll something plausible but wrong and you'll say, "Oh, chat GPT is hallucinating say, "Oh, chat GPT is hallucinating say, "Oh, chat GPT is hallucinating again." Maybe it'll turn an assumption again." Maybe it'll turn an assumption again." Maybe it'll turn an assumption into a recommendation. And because the into a recommendation. And because the into a recommendation. And because the output looks really clean, people will output looks really clean, people will output looks really clean, people will stop noticing where it came from. And stop noticing where it came from. And stop noticing where it came from. And that's the risk. It's not evil AI. It's that's the risk. It's not evil AI. It's that's the risk. It's not evil AI. It's that unowned work starts to have real that unowned work starts to have real that unowned work starts to have real consequences over time because people consequences over time because people consequences over time because people don't check it and care for it. So what don't check it and care for it. So what don't check it and care for it. So what do you do with your agent? I would start do you do with your agent? I would start do you do with your agent? I would start with four things. And I want to keep with four things. And I want to keep with four things. And I want to keep this really, really simple. Give it a this really, really simple. Give it a this really, really simple. Give it a job. Give it a diet. Give it boundaries.

job. Give it a diet. Give it boundaries. job. Give it a diet. Give it boundaries. And give it a review loop. The job is And give it a review loop. The job is And give it a review loop. The job is what this agent is supposed to do. Not what this agent is supposed to do. Not what this agent is supposed to do. Not help with the product. Not something help with the product. Not something help with the product. Not something vague, right? Not make me more vague, right? Not make me more vague, right? Not make me more productive. a real job that matters, productive. a real job that matters, productive. a real job that matters, like prepare first pass backlog items like prepare first pass backlog items like prepare first pass backlog items for refinement, draft refund replies for for refinement, draft refund replies for for refinement, draft refund replies for this ticket type, inspect a poll request this ticket type, inspect a poll request this ticket type, inspect a poll request for risky changes, build me a weekly for risky changes, build me a weekly for risky changes, build me a weekly research brief from these sources. If research brief from these sources. If research brief from these sources. If you can't say the job in a sentence, the you can't say the job in a sentence, the you can't say the job in a sentence, the agent is probably too vague. The diet is agent is probably too vague. The diet is agent is probably too vague. The diet is what the agent reads. This matters a lot what the agent reads. This matters a lot what the agent reads. This matters a lot more than people think. Agents eat more than people think. Agents eat more than people think. Agents eat context, right? They eat docs and context, right? They eat docs and context, right? They eat docs and tickets and transcripts and repo tickets and transcripts and repo tickets and transcripts and repo instructions and examples and whatever instructions and examples and whatever instructions and examples and whatever else you put in front of them. If the else you put in front of them. If the else you put in front of them. If the diet is stale or bloated, the agent can diet is stale or bloated, the agent can diet is stale or bloated, the agent can get stale and bloated. If the diet is get stale and bloated. If the diet is get stale and bloated. If the diet is messy, the agent can get messy. If the messy, the agent can get messy. If the messy, the agent can get messy. If the diet uses incorrect examples, the agent diet uses incorrect examples, the agent diet uses incorrect examples, the agent learns bad habits. This is why the learns bad habits. This is why the learns bad habits. This is why the little Pokemon analogy works so well for little Pokemon analogy works so well for little Pokemon analogy works so well for me. Collecting the Pokémons isn't the me. Collecting the Pokémons isn't the me. Collecting the Pokémons isn't the point. You have to know what each point. You have to know what each point. You have to know what each Pokemon is good at. You have to know Pokemon is good at. You have to know Pokemon is good at. You have to know where not to use it. You have to know where not to use it. You have to know where not to use it. You have to know what it has been trained on, and you what it has been trained on, and you what it has been trained on, and you have to notice when it starts picking up have to notice when it starts picking up have to notice when it starts picking up bad habits. And that sounds kind of bad habits. And that sounds kind of bad habits. And that sounds kind of silly, but the responsibility for care silly, but the responsibility for care silly, but the responsibility for care is real. And then come the boundaries.

is real. And then come the boundaries. is real. And then come the boundaries. What can this agent touch? Can it read What can this agent touch? Can it read What can this agent touch? Can it read files? Can it draft? Can it write to the files? Can it draft? Can it write to the files? Can it draft? Can it write to the system? Can it send or delete? Can it system? Can it send or delete? Can it system? Can it send or delete? Can it update Jura? These are not the same update Jura? These are not the same update Jura? These are not the same level of risk. A drafton agent is one level of risk. A drafton agent is one level of risk. A drafton agent is one thing. An agent that can write into a thing. An agent that can write into a thing. An agent that can write into a system of record is much much more system of record is much much more system of record is much much more serious. An agent that can send a serious. An agent that can send a serious. An agent that can send a customer message or merge code is in a customer message or merge code is in a customer message or merge code is in a different category entirely. And your different category entirely. And your different category entirely. And your sense of ownership should move with sense of ownership should move with sense of ownership should move with that. You should feel emotional stakes. that. You should feel emotional stakes. that. You should feel emotional stakes. So, the rule is simple. Start with read So, the rule is simple. Start with read So, the rule is simple. Start with read only. Start with draft only if you're only. Start with draft only if you're only. Start with draft only if you're unsure. Let the agent earn more unsure. Let the agent earn more unsure. Let the agent earn more permission and make sure you're permission and make sure you're permission and make sure you're comfortable moving outside that narrow comfortable moving outside that narrow comfortable moving outside that narrow job. And then the review loop is how you job. And then the review loop is how you job. And then the review loop is how you keep it healthy. And this is one of the keep it healthy. And this is one of the keep it healthy. And this is one of the phrases people like to use. Loop system phrases people like to use. Loop system phrases people like to use. Loop system workflow. It's getting used a lot and I workflow. It's getting used a lot and I workflow. It's getting used a lot and I think it can sound more complicated than think it can sound more complicated than think it can sound more complicated than it is. A loop just means the work comes it is. A loop just means the work comes it is. A loop just means the work comes back around. The agent runs a human back around. The agent runs a human back around. The agent runs a human reviews. Maybe another agent reviews. In reviews. Maybe another agent reviews. In reviews. Maybe another agent reviews. In some cases, mostly a human reviews. You some cases, mostly a human reviews. You some cases, mostly a human reviews. You notice what was good and what was bad. notice what was good and what was bad. notice what was good and what was bad. You update the instructions, the You update the instructions, the You update the instructions, the sources, the permission, the agent runs sources, the permission, the agent runs sources, the permission, the agent runs again. There's the loop. It's not magic. again. There's the loop. It's not magic. again. There's the loop. It's not magic. It's not a giant governance process. It's not a giant governance process. It's not a giant governance process. It's not the best thing since sliced It's not the best thing since sliced It's not the best thing since sliced bread. It's just a way work gets done bread. It's just a way work gets done bread. It's just a way work gets done with agents. Run, review, improve, run with agents. Run, review, improve, run with agents. Run, review, improve, run again. So, let me give you a product again. So, let me give you a product again. So, let me give you a product team version because this is where this team version because this is where this team version because this is where this becomes really obvious with a specific becomes really obvious with a specific becomes really obvious with a specific example. Imagine a scrum team building a example. Imagine a scrum team building a example. Imagine a scrum team building a new onboarding flow. Every week before new onboarding flow. Every week before new onboarding flow. Every week before backlog refinement, someone has to do backlog refinement, someone has to do backlog refinement, someone has to do the same messy prep. You read the the same messy prep. You read the the same messy prep. You read the customer tickets. You check the PRD. You customer tickets. You check the PRD. You customer tickets. You check the PRD. You look at the design changes. You review look at the design changes. You review look at the design changes. You review the backlog. You pull the pain and you the backlog. You pull the pain and you the backlog. You pull the pain and you turn it into acceptance criteria and turn it into acceptance criteria and turn it into acceptance criteria and dependencies and tickets. It's real dependencies and tickets. It's real dependencies and tickets. It's real work. It's a lot of product managers work. It's a lot of product managers work. It's a lot of product managers weeks. So the PM starts to build a story

weeks. So the PM starts to build a story weeks. So the PM starts to build a story prep agent to help themselves. And the prep agent to help themselves. And the prep agent to help themselves. And the first vision is simple. It's going to first vision is simple. It's going to first vision is simple. It's going to read the current PRD, the design brief, read the current PRD, the design brief, read the current PRD, the design brief, the tag support tickets, the backlog, the tag support tickets, the backlog, the tag support tickets, the backlog, and a few examples of good stories. And and a few examples of good stories. And and a few examples of good stories. And then it's going to prepare a refinement then it's going to prepare a refinement then it's going to prepare a refinement packet. Not final tickets, just a packet. Not final tickets, just a packet. Not final tickets, just a packet. It's going to say, "Here are the packet. It's going to say, "Here are the packet. It's going to say, "Here are the candidate stories. Here's the customer candidate stories. Here's the customer candidate stories. Here's the customer evidence. Here are the acceptance evidence. Here are the acceptance evidence. Here are the acceptance criteria. Here are the dependencies. criteria. Here are the dependencies. criteria. Here are the dependencies. here are the assumptions I'm making and here are the assumptions I'm making and here are the assumptions I'm making and here's the open decisions my human needs here's the open decisions my human needs here's the open decisions my human needs to make. It's a great agent job to start to make. It's a great agent job to start to make. It's a great agent job to start with. You can do that in codeex or with. You can do that in codeex or with. You can do that in codeex or claude, but now imagine the team starts claude, but now imagine the team starts claude, but now imagine the team starts relying on that packet every week. It's relying on that packet every week. It's relying on that packet every week. It's now a team agent. Now the agent is now a team agent. Now the agent is now a team agent. Now the agent is shaping the sprint. If the PRD is old, shaping the sprint. If the PRD is old, shaping the sprint. If the PRD is old, old product assumptions are going to old product assumptions are going to old product assumptions are going to enter real work. If the support tickets enter real work. If the support tickets enter real work. If the support tickets are noisy, agent oversight start to are noisy, agent oversight start to are noisy, agent oversight start to matter. If the design changed yesterday matter. If the design changed yesterday matter. If the design changed yesterday and the agent didn't pick up on that and the agent didn't pick up on that and the agent didn't pick up on that because of a job conflict, that's going because of a job conflict, that's going because of a job conflict, that's going to matter. It's not going to be an to matter. It's not going to be an to matter. It's not going to be an explosion or a robot takeover, right? explosion or a robot takeover, right? explosion or a robot takeover, right? but it's going to be an issue for the but it's going to be an issue for the but it's going to be an issue for the team. So what does ownership look like team. So what does ownership look like team. So what does ownership look like in that context? The product manager in that context? The product manager in that context? The product manager owns the job because the product manager owns the job because the product manager owns the job because the product manager owns backlog quality. The operating team owns backlog quality. The operating team owns backlog quality. The operating team has to own the operating agent. And the has to own the operating agent. And the has to own the operating agent. And the maintenance loop is not complicated. So maintenance loop is not complicated. So maintenance loop is not complicated. So what does ownership look like? The what does ownership look like? The what does ownership look like? The product manager owns the job and is the product manager owns the job and is the product manager owns the job and is the single threaded owner that should care single threaded owner that should care single threaded owner that should care about whether the agent works or not.

about whether the agent works or not. about whether the agent works or not. That simple. The engineering lead can That simple. The engineering lead can That simple. The engineering lead can help with technical assumptions. The QA help with technical assumptions. The QA help with technical assumptions. The QA team can help with testability. The AI team can help with testability. The AI team can help with testability. The AI team might help with tooling in some team might help with tooling in some team might help with tooling in some cases and the maintenance loop for that cases and the maintenance loop for that cases and the maintenance loop for that agent, it's not as complicated as it agent, it's not as complicated as it agent, it's not as complicated as it might sound, even though it's an agent might sound, even though it's an agent might sound, even though it's an agent that does a lot of work. Before that does a lot of work. Before that does a lot of work. Before refinement, the agent prepares the refinement, the agent prepares the refinement, the agent prepares the packet and the PM should review that. packet and the PM should review that. packet and the PM should review that. During refinement, the team can notice During refinement, the team can notice During refinement, the team can notice where it helped and where it confused where it helped and where it confused where it helped and where it confused the discussion. After the sprint, the the discussion. After the sprint, the the discussion. After the sprint, the owner, the PM should check a few owner, the PM should check a few owner, the PM should check a few stories. Did engineers have to rewrite stories. Did engineers have to rewrite stories. Did engineers have to rewrite them? Did QA understand them? Did them? Did QA understand them? Did them? Did QA understand them? Did dependencies show up late? If you can dependencies show up late? If you can dependencies show up late? If you can fix the inputs to the agent and change fix the inputs to the agent and change fix the inputs to the agent and change your output, you can fix the system. So your output, you can fix the system. So your output, you can fix the system. So remove the stale PRD, add a better remove the stale PRD, add a better remove the stale PRD, add a better example, change what it is allowed to example, change what it is allowed to example, change what it is allowed to read. That's care and feeding. And read. That's care and feeding. And read. That's care and feeding. And that's how you go beyond prompting and that's how you go beyond prompting and that's how you go beyond prompting and start to work like you're in 2026 with start to work like you're in 2026 with start to work like you're in 2026 with agents and loops. Prompting is asking, agents and loops. Prompting is asking, agents and loops. Prompting is asking, agent work is giving context and care agent work is giving context and care agent work is giving context and care and feeding to the agent so it can do and feeding to the agent so it can do and feeding to the agent so it can do its job. There's a big difference its job. There's a big difference its job. There's a big difference between asking write acceptance criteria between asking write acceptance criteria between asking write acceptance criteria for this feature, please, and giving an for this feature, please, and giving an for this feature, please, and giving an agent a job. The job might look like agent a job. The job might look like agent a job. The job might look like read the PRD, the last 20 support read the PRD, the last 20 support read the PRD, the last 20 support tickets, the design brief, and our three tickets, the design brief, and our three tickets, the design brief, and our three best backlog examples. Draft the stories best backlog examples. Draft the stories best backlog examples. Draft the stories for refinement, attach the customer for refinement, attach the customer for refinement, attach the customer evidence, mark assumptions, and don't evidence, mark assumptions, and don't evidence, mark assumptions, and don't create juror tickets. Put everything create juror tickets. Put everything create juror tickets. Put everything into review so I can look at it first.

into review so I can look at it first. into review so I can look at it first. Do you feel that difference? One is a Do you feel that difference? One is a Do you feel that difference? One is a prompt, and the PM may feel more prompt, and the PM may feel more prompt, and the PM may feel more productive, but it may not hit the team. productive, but it may not hit the team. productive, but it may not hit the team. The other is a job with sources and The other is a job with sources and The other is a job with sources and boundaries and output and a review loop, boundaries and output and a review loop, boundaries and output and a review loop, and it absolutely affects the entire and it absolutely affects the entire and it absolutely affects the entire team. And this is the move that most team. And this is the move that most team. And this is the move that most people need to make in 2026. From people need to make in 2026. From people need to make in 2026. From prompts to jobs. Now, if you're thinking prompts to jobs. Now, if you're thinking prompts to jobs. Now, if you're thinking about this as a team leader, the about this as a team leader, the about this as a team leader, the question becomes bigger, but it doesn't question becomes bigger, but it doesn't question becomes bigger, but it doesn't need to become more abstract. This is a need to become more abstract. This is a need to become more abstract. This is a danger spot where agents can go unowned. danger spot where agents can go unowned. danger spot where agents can go unowned. Nobody owns the AR agent that summarizes Nobody owns the AR agent that summarizes Nobody owns the AR agent that summarizes performance notes before calibration. performance notes before calibration. performance notes before calibration. So, nobody checks whether it's pulling So, nobody checks whether it's pulling So, nobody checks whether it's pulling from stale manager feedback or from stale manager feedback or from stale manager feedback or flattening important context. That's a flattening important context. That's a flattening important context. That's a real example, by the way. Nobody owns real example, by the way. Nobody owns real example, by the way. Nobody owns the recruiting agent that drafts the recruiting agent that drafts the recruiting agent that drafts candidate scorecards. So, no one's candidate scorecards. So, no one's candidate scorecards. So, no one's accountable when it drifts. No one knows accountable when it drifts. No one knows accountable when it drifts. No one knows the support triage agent. So refund the support triage agent. So refund the support triage agent. So refund policy may be stale and may be policy may be stale and may be policy may be stale and may be misapplied. You see the same thing misapplied. You see the same thing misapplied. You see the same thing cropping up all over the business when cropping up all over the business when cropping up all over the business when you have projects that parachute in from you have projects that parachute in from you have projects that parachute in from an AI team and that end up unowned in an AI team and that end up unowned in an AI team and that end up unowned in those target teams. Finance can have the those target teams. Finance can have the those target teams. Finance can have the same thing, right? You're going to need same thing, right? You're going to need same thing, right? You're going to need an agent roster as a team lead. Not a an agent roster as a team lead. Not a an agent roster as a team lead. Not a big database, right? Just a list of big database, right? Just a list of big database, right? Just a list of agents your team is using. This is our agents your team is using. This is our agents your team is using. This is our story prep agent. This is our release story prep agent. This is our release story prep agent. This is our release note agent. This is our customer call note agent. This is our customer call note agent. This is our customer call summary agent. This is our PR review summary agent. This is our PR review summary agent. This is our PR review agent. And for each of them, you should agent. And for each of them, you should agent. And for each of them, you should know who the owner is. You should know know who the owner is. You should know know who the owner is. You should know who what the sources are. The agent can who what the sources are. The agent can who what the sources are. The agent can look at what the permissions are, the look at what the permissions are, the look at what the permissions are, the review cadence, and the known failure review cadence, and the known failure review cadence, and the known failure modes. And that's it. Because once the modes. And that's it. Because once the modes. And that's it. Because once the agent is visible, you can manage it. As agent is visible, you can manage it. As agent is visible, you can manage it. As a team leader, you can manage it. If a team leader, you can manage it. If a team leader, you can manage it. If it's invisible, it just becomes this it's invisible, it just becomes this it's invisible, it just becomes this weird shadow process where work is weird shadow process where work is weird shadow process where work is moving through tools and nobody can

moving through tools and nobody can moving through tools and nobody can explain how the output got there. And explain how the output got there. And explain how the output got there. And there's these anecdotes that people will there's these anecdotes that people will there's these anecdotes that people will bring up in performance reviews that bring up in performance reviews that bring up in performance reviews that say, "I'm AI native." Now, that's not say, "I'm AI native." Now, that's not say, "I'm AI native." Now, that's not productive. And I get it. The energy productive. And I get it. The energy productive. And I get it. The energy right now is build, build, build build. right now is build, build, build build. right now is build, build, build build. But if every ambitious person in the But if every ambitious person in the But if every ambitious person in the company creates three agents, you don't company creates three agents, you don't company creates three agents, you don't necessarily have more productivity if necessarily have more productivity if necessarily have more productivity if you don't scale that ownership. I love you don't scale that ownership. I love you don't scale that ownership. I love the build energy. It's not bad. It's the build energy. It's not bad. It's the build energy. It's not bad. It's powerful, but powerful things need powerful, but powerful things need powerful, but powerful things need owners. And if you want the simplest owners. And if you want the simplest owners. And if you want the simplest version of this, here is the owner card version of this, here is the owner card version of this, here is the owner card that you should take with you. For every that you should take with you. For every that you should take with you. For every agent that matters, write down the name, agent that matters, write down the name, agent that matters, write down the name, the owner, the job, the sources, what it the owner, the job, the sources, what it the owner, the job, the sources, what it can do, what it can't do, and the can do, what it can't do, and the can do, what it can't do, and the failure mode you need to watch for. I failure mode you need to watch for. I failure mode you need to watch for. I said that was a spreadsheet for team said that was a spreadsheet for team said that was a spreadsheet for team leaders, but that card can also be the leaders, but that card can also be the leaders, but that card can also be the agent ownership card for individuals. agent ownership card for individuals. agent ownership card for individuals. Literally, you can make a Slack channel Literally, you can make a Slack channel Literally, you can make a Slack channel with a bunch of agent owner cards where with a bunch of agent owner cards where with a bunch of agent owner cards where you can say, "This is my agent. This is you can say, "This is my agent. This is you can say, "This is my agent. This is what it does." And like people can share what it does." And like people can share what it does." And like people can share them. And this is something where if them. And this is something where if them. And this is something where if you're trying to create agentto agent you're trying to create agentto agent you're trying to create agentto agent collaboration in the company, which a collaboration in the company, which a collaboration in the company, which a lot of people are doing through Slack, lot of people are doing through Slack, lot of people are doing through Slack, you kind of need a way for the humans to you kind of need a way for the humans to you kind of need a way for the humans to understand what these agents are doing understand what these agents are doing understand what these agents are doing at the top and to have almost an agent at the top and to have almost an agent at the top and to have almost an agent registry so the humans can understand registry so the humans can understand registry so the humans can understand what's going on. And yes, this is on what's going on. And yes, this is on what's going on. And yes, this is on purpose a little bit like Google's ATA purpose a little bit like Google's ATA purpose a little bit like Google's ATA protocol. Google's ATA protocol assumes protocol. Google's ATA protocol assumes protocol. Google's ATA protocol assumes that agents need introduction cards like that agents need introduction cards like that agents need introduction cards like that to each other. That's great, but that to each other. That's great, but that to each other. That's great, but what about the humans? We need a sense what about the humans? We need a sense what about the humans? We need a sense of ownership almost like a certificate of ownership almost like a certificate of ownership almost like a certificate for the agent. And I find that companies for the agent. And I find that companies for the agent. And I find that companies that have that mindset, regardless of

that have that mindset, regardless of that have that mindset, regardless of what they call it, do better because what they call it, do better because what they call it, do better because they take agent ownership more they take agent ownership more they take agent ownership more seriously. And they should than agent seriously. And they should than agent seriously. And they should than agent building. Just building a new agent, you building. Just building a new agent, you building. Just building a new agent, you shouldn't get credit for these days. shouldn't get credit for these days. shouldn't get credit for these days. Owning an agent and using it to deliver Owning an agent and using it to deliver Owning an agent and using it to deliver value, that's what you should get credit value, that's what you should get credit value, that's what you should get credit for. And I think this is what catching for. And I think this is what catching for. And I think this is what catching up with AI actually looks like today. up with AI actually looks like today. up with AI actually looks like today. It's not having the most agents. It's It's not having the most agents. It's It's not having the most agents. It's not knowing every tool. It's not winning not knowing every tool. It's not winning not knowing every tool. It's not winning a vocabulary argument about loops and a vocabulary argument about loops and a vocabulary argument about loops and evals. It's having a small number of evals. It's having a small number of evals. It's having a small number of agents that you own and that deliver agents that you own and that deliver agents that you own and that deliver real value in workflows. You know what real value in workflows. You know what real value in workflows. You know what they do. You know what they eat. You they do. You know what they eat. You they do. You know what they eat. You know what they can touch. You know how know what they can touch. You know how know what they can touch. You know how you review them. You know when to trust you review them. You know when to trust you review them. You know when to trust them and when not to do. They're your them and when not to do. They're your them and when not to do. They're your pets. You know what work you delegated pets. You know what work you delegated pets. You know what work you delegated and what responsibility you kept. This and what responsibility you kept. This and what responsibility you kept. This is the grown-up version of Asian is the grown-up version of Asian is the grown-up version of Asian adoption. Ironically, as a child of the adoption. Ironically, as a child of the adoption. Ironically, as a child of the 80s and 90s, Pokemon prepared me well. 80s and 90s, Pokemon prepared me well. 80s and 90s, Pokemon prepared me well. Prompting was the first skill for us, Prompting was the first skill for us, Prompting was the first skill for us, right? Back in 2023, we had to learn how right? Back in 2023, we had to learn how right? Back in 2023, we had to learn how to ask better questions and and that's to ask better questions and and that's to ask better questions and and that's still a useful skill because it forces still a useful skill because it forces still a useful skill because it forces us to articulate. Delegation was the us to articulate. Delegation was the us to articulate. Delegation was the next skill. We had to learn how to hand next skill. We had to learn how to hand next skill. We had to learn how to hand over real work. That's a very 2025 over real work. That's a very 2025 over real work. That's a very 2025 thing. Maintenance is a 2026 skill thing. Maintenance is a 2026 skill thing. Maintenance is a 2026 skill because useful agents become our because useful agents become our because useful agents become our responsibility. So, here's your decision responsibility. So, here's your decision responsibility. So, here's your decision rule. Very easy. If a system can read rule. Very easy. If a system can read rule. Very easy. If a system can read important context, produce work you act important context, produce work you act important context, produce work you act on or your team acts on, touch a on or your team acts on, touch a on or your team acts on, touch a workflow other people depend on, it workflow other people depend on, it workflow other people depend on, it needs an owner. Now, if it's yours, you needs an owner. Now, if it's yours, you needs an owner. Now, if it's yours, you own it. If it belongs to the team, the own it. If it belongs to the team, the own it. If it belongs to the team, the team needs to name one person as an team needs to name one person as an team needs to name one person as an owner. And if nobody's willing to own

owner. And if nobody's willing to own owner. And if nobody's willing to own it, it probably shouldn't be doing it, it probably shouldn't be doing it, it probably shouldn't be doing important work, and you should think important work, and you should think important work, and you should think about decommissioning it. about decommissioning it. about decommissioning it. Look, I put the deeper checklist and the Look, I put the deeper checklist and the Look, I put the deeper checklist and the owner card and the whole guide for how owner card and the whole guide for how owner card and the whole guide for how to sort of develop an agent registry at to sort of develop an agent registry at to sort of develop an agent registry at the company or frankly for you, your the company or frankly for you, your the company or frankly for you, your Pokemon collection agent set over on Pokemon collection agent set over on Pokemon collection agent set over on Substack because that works much better Substack because that works much better Substack because that works much better as a written guide. But the mental model as a written guide. But the mental model as a written guide. But the mental model really is this simple. Stop asking only really is this simple. Stop asking only really is this simple. Stop asking only whether you can build an agent. Start whether you can build an agent. Start whether you can build an agent. Start asking whether you can care and feed it. asking whether you can care and feed it. asking whether you can care and feed it. Every agent needs an owner. Not because Every agent needs an owner. Not because Every agent needs an owner. Not because agents are bad, because useful agents agents are bad, because useful agents agents are bad, because useful agents must become part of our work in 2026. must become part of our work in 2026. must become part of our work in 2026. That is the bar. So, take care of your That is the bar. So, take care of your That is the bar. So, take care of your little Pokemon agents. I hope this has little Pokemon agents. I hope this has little Pokemon agents. I hope this has helped clear up a lot of the buzz and helped clear up a lot of the buzz and helped clear up a lot of the buzz and the confusion around what an agent is the confusion around what an agent is the confusion around what an agent is and what our job is with agents in 2026. and what our job is with agents in 2026. and what our job is with agents in 2026. I'll see you next time.