## Intro [00:00] Hey everybody, I'm Jamie. Welcome back to  Teachers Tech. So, in my last Claude Code video, we did something pretty cool. We built a full  bookmark dashboard app just by typing in plain English. No coding experience needed. If you  haven't seen that one, I'll link it right up here and down below in the description. Definitely  go watch that one first because today's video builds on everything we covered there. But today,  we're going to go way beyond building apps. Today, we're going to be building an AI agent. Let  me show you what I mean. Are you seeing this? I typed one sentence and it started thinking on  its own. It broke the task into steps. It asked me follow-up questions. It went out to gather  information, organized everything, and produced a full report all by itself. Last time, Cloud  Code did what we told it to do. This time, we're going to teach it to figure out things on its own.  That's called an agentic workflow, and honestly, it's the most useful thing I've learned in AI this  year. Oh, and we're doing the whole thing inside VS Code this time. way more comfortable than  the raw terminal like last time. Let's get into ## What is an agentic workflow [01:00] it. All right, before we start building anything,  let's talk about what an agentic workflow actually is because it sounds way more complicated than it  really is. Think back to the last video. We told Claude Code exactly what to build. We said, "Make  me a bookmark dashboard with search dark mode." We gave it specific instructions and it followed  them. That was powerful, but we're still doing all the thinking. We were directing every step.  An agentic workflow flips that around. Let me break it down into three levels so you can see the  differences. Level one is chat. This is the most basic way people use AI. You ask a question, you  get an answer. One and done. That's your standard chat GPT or clawed conversation. Most people never  get past this level. Level two is building. This is what we did in the last video. You tell  the AI what you want to create. You guide the process step by step and it writes the code for  you. You're the director, the AI is the builder. It's way more powerful than just chatting, but  you're still making all the decisions. Level three is agentic. And this is where things get really  interesting. Instead of telling the AI what to do step by step, you describe the goal. You say,  "Here's what I need to accomplish." And the AI figures out the steps on its own. It makes a plan,  picks its tools, adapts when things don't go as expected, and it will even ask you questions when  it's not sure about something. That's what we're doing today. Here's the way I think about it. It's  like the difference between micromanaging a brand new employee versus trusting an experienced one.  In the last video, we were the manager saying, "Do this, then this, then this." Today,  we're saying, "Here's the goal. Here's what you have access to. Figure it out." That's a  completely different relationship with the AI. So what specifically makes something agentic? Three  things. First, it follows a process. It's not just answering a single question. It's working through  a whole series of steps to get a result. Second, it makes decisions. It chooses what to do based  on what it finds along the way. If one approach isn't working, it pivots. If it finds something  unexpected, it adjusts. Third, it asks before it assumes. Instead of guessing what you want, a good  aenic workflow will ask you clarifying questions first. That one change alone makes the output  10 times better. And the use cases for this are endless. You could use it to research a topic  across multiple sources and compile a report. Analyze a folder of files and organize them by  category. Draft personalize emails for a whole list of contacts. Compare competitors and produce  a sidebyside breakdown. The pattern is always the same. You give it the goal and it does the  work. All right, let's get our tool set up. Now, ## Setting up VS Code with Claude Code [03:37] last time we worked entirely in the terminal and  that works, but VS Code gives us a much better experience for what we're going to do today. You  can see your files, your code, and cloud code all in the same place. When the agent is creating and  editing multiple files, which it will be, having that visual file explorer makes a huge difference.  So, let's get this set up. This only takes about 2 minutes. If you don't have Visual Studio Code  on your computer, I'll put the link to this site down below in the description. Then you can pick  what operating system you're working on. Today I'm going to be on Windows, so I'm going to go ahead  and choose this and download this for my machine. Once you have it downloaded, go ahead open it up.  Where I want you to go is right here to extensions and you're going to type in just claude code  up here. And you're going to see that we can install cloud code here. So I'm just going to go  ahead and click on this. I'm going to trust the publisher and this install. All right, it's all  ready to go. We have it installed. I can see the cloud code. I'm going to click on it to open. I'm  going to go and mark this as done. I'm just going to go and close these extensions over here. I'll  just click on it again to give myself more room. Notice you can uh stretch these uh to whatever  you want that works best for your preferences. If you already installed Cloud Code from the  last video, you're all set. The extension uses the same installation. If you haven't installed it  yet, go watch that first video because we covered the whole setup there. Let me give you a quick  orientation of VS Code. Over here on the left, ## Quick orientation of VS Code [05:07] we've got our file explorer. This is where you're  going to see every file the agent creates in real time. Folders appearing, documents being written.  It's really satisfying to watch. Over here, we have our cloud code panel. This is where you talk  to the agent. You go down here, type your prompts, you see what it's thinking, and you watch it work.  Now, let's create our project folder. I'm going to go ahead and create a new folder where we're going  to be working out of. I'm just going to click open folder. And I haven't created one yet. If you know  what folder you want, you can go ahead and choose that one. I'm going to go in my documents here.  And I'm just going to create a brand new folder. And I'm going to go call it my first agent.  Select folder. I'm just going to go back and click on clawed code here and new session just to  open that back up. And if I click on the explorer, you can see here's the folder that I just created.  Now, there's a couple of features in VS Code that are going to be really important for you today.  First, you can see the files updating real time as the agent works. Watch the file explorer. When the  agent creates something, it just appears. Second, when the agent plans a multi-step task, you'll  see a to-do list pop up showing you exactly what's going to do and where in the process.  And here's the big one, planning mode. Press shift plus tab and you'll toggle into planning  mode. In this mode, Claude Code thinks and plans but doesn't actually change any files. It's like  a dry run. We're going to use this a lot today because the key to aenic workflows is planning  before building. All right, our environment is ready. Now, let's set up one more thing that's  really going to make all the difference. Okay, this is something I didn't cover in the first  video and it completely changes how claude code ## Setting up your CLAUDE.md file [06:42] works for you. It's a file called claude.md.  And I'm not exaggerating when I say this is the single most important thing you can set up.  So, what is claw.md? It's a simple markdown file that you put in your project folder. Claude code  reads it automatically every time it starts up. Think of it like an onboarding document for a  new employee. It tells the agent who you are, what the project's about, and how you want things  done. Let's build one together right now. In VS Code, I'm creating a new file in the root of our  project. I'll name it claude.md all caps. And now, let's fill it in section by section. Okay, the  first section is going to be project context. And this is going to tell the agent what the workspace  is for. So I am going to give it a heading right here at the beginning. And to do this, use the  pound key and then a space. And then you can just type your heading. You'll notice how it turns  blue. I'm just going to give it a couple spaces just so everything's easier to read and so I can  take a quick glance at and see what's happening. I'm going to go and put this. So for project  context, this is my AI agent workspace. I use it for research, content creation, and productivity  workflows. I'm going to give it a couple spaces. Next is the about me. This is where you tell the  agent about who you are, what what you care about. So, I'll give it again another header and give  it a couple spaces. And this is what I'm going to put. I create content about technology and  productivity. My audience is people who want practical, nononsense tutorials. I prefer clear,  jargon-free output. Now, the most important part, rules. This is where you set the guardrails. These  are the instructions the agent will follow every single time. Let's give it a heading and a couple  spaces. And my rules are going to be listed with dashes in front of them. So you can have a clear  list just like this. So always ask clarifying questions before starting a complex task. Show  your plans and steps before executing. Keep reports and summaries concise, bullet points over  paragraphs. Save all output files to the output folder and site sources when doing research. And  finally, project structure. This tells the agent where things go. And this is where I want things.  I have my workflows. Workflow instruction files. My output are the finished deliverables. And  resources are the reference docs and templates. I'm just going to save this now by hitting ctrls  together. That took us about 2 minutes. But now, every single time you open Claude Code in this  folder, it already knows your preferences, your rules, and exactly how you want things organized.  No repeating yourself. Let me show you why this matters so much. Now, I'm going to use the same  prompt, research remote work trends on the left ## CLAUDE.md comparison -- with vs without [09:22] without claw.md. I made a separate folder with no  claw.md file in it. So, I get this generic wall of text. You can see as it goes through, there's  not really much structure. I get one question just about allowing something uh just kind of a data  dump. But on the right with the claw.md file, the agent's asking me many different questions  as it goes through it. Then it gives me organized bullet list, clear sections, source cited, and  save the file exactly where I told it to. Let's take a look at this end file here. Here's the link  to the report, and it's in the output. Remember, I told it to uh go ahead and put it the finished  output in here. So, if I click on it, it brings me to this. Now, you might be thinking, well, that  doesn't look very good. But there's a shortcut to see what it looks like. So, if you go to control,  shift, and v uh together, when you click in here, if you press all those keys together, you get  the report, what it created for you. So, as I go through it, you can see how much more detailed  uh this was. Here we have our key takeaways. we have all of this information versus that that one  little short link that we had with the uh without the claw.md file. And this is why the claw.md is  non-negotiable. 2 minutes of setup saves you hours of frustration down the road. And here's something  that was just added to claude code that makes this ## Automatic memory in Claude Code [10:50] even more powerful. Automatic memory. As the agent  works through your project, it now automatically records things it learns and recalls them in the  future sessions. So, it's just not reading your claw.md at startup anymore. It's also building its  own understanding of your project over time. The more you use it in this folder, the smarter it  gets about your preferences and how things are organized. Before we start building our agent,  I want to give you the mental model. There are three pieces that make agendic workflows work. And  once you understand them, everything else clicks ## The three layers of agentic workflows [11:16] into place. Layer one, workflows. These are just  plain English files that describe a process step by step. Think of them like a recipe. The agent  reads the recipe and follows it, but it's smart enough to adapt if anything's off. Workflows live  as a simple markdown file in your project. Nothing fancy. Layer two, the agent. This is just  cloud code itself. It reads your workflows, thinks through the steps, and makes decisions. You  don't program the agent. You don't write the code for it. You just give it clear instructions and  let it reason. Layer three tools. These are what the agent uses to actually get things done. Out  of the box, Claude Code can read and write files, run terminal commands, and search the web. That's  right. Web search is built right in. No extra setup needed. The agent can go out, find current  information, and bring it back into your project. Now, you can also connect external services like  Gmail, Notion, and databases using something called MCP. That's model context protocol. That  opens up a whole world of possibilities, and we'll cover that in depth in the next video. But for  today, the built-in tools are more than enough for everything we're doing. And here's the key insight  that most people miss. Most people skip straight to the fancy tools and plugins. But the real power  is writing good workflows. A well-written workflow with basic tools will outperform a sloppy  prompt with every plug-in in the world. The workflow is where the magic lives. Okay, enough  theory. Let's build our first agent. All right, ## Building your first agent with plan mode [12:45] I'm in our empty project. You can see I just have  the claw.md file that we created. I deleted those other test folders when I ran through that quick  test. The uh first thing we're going to do is go and change it to plan mode. So, when I'm inside  the chat down here, I'm just going to press shift and tab. But notice if I press it once, it goes  to edit automatically. I need to press it one more time, it brings me to plan mode. If I press  it again, it brings me back to where I was. So, two times pressing shift and tap brings me over  to plan mode. Now, plan mode means Cloud Code will think and plan, but won't change any files. We're  going to design our workflow before building it. Always plan first. Now, I'm going to describe what  I want. And watch carefully. I'm not giving it step-by-step instructions. I'm describing my goal.  And this is what I'm going to say. I want to build a workflow where I can give you a topic. You'll  research it thoroughly, organize the findings, and produce a clean, structured report. Before you  start researching, you should ask me clarifying questions about the scope, the audience, and what  I specifically want to know. I'm going to submit. Look at this. I gave it a goal and it's already  breaking it down into steps. It's thinking about what the workflow needs. A clarifying question  phase, a research phase, a synthesis phase, an output phase. It's even suggesting quality  checks. That's the agendic thinking and action. I didn't tell it these steps. It figured them out.  But I want to add one more thing before we build. Before I add one more thing, let's go ahead and  take a look what it has already. Here we have a resource plan workflow. And if I scroll through,  I can see there's some files that are going to be created. And remember once we accept it to keep an  eye over on explore here because you're going to see it get created. It's going to uh go ahead  and create this research report uh.md as I go down. Step two, confirm the plan. Three, research.  Four, write the report. Five, save the output. It gave me some formatting. So, if I like everything  here, I could go ahead yes and auto accept or yes, manually approve or keep planning. But what I'm  going to say is this. I also want the report to include a section at the end with key takeaways  and recommendations for the next step. So, I'm just going to put that in and hit enter. So,  it's going to update the plan. It doesn't just start over. It just folds the new requirement into  what it already had. So, that's why plan mode is so valuable. you can shape the approach before  anything gets built. So I can see now it says uh right here plan updated the report template  will now include two additional sections here I have recommended next steps key takeaways. So I'm  going to go ahead and you know what I'm going to I'm happy with it. Let's build it and I'm going  to go to auto accept. So it's just going to go through and accept everything. So you're going  to see the mode change from plan and as I click on this just keep an eye on explorer. So notice  we have our workflows. If I go ahead and open it, it's still working, but we have the research  report MD that I said it was going to create. Now let's look at what it actually wrote. This is  the important part. So I'm I'm going to go to the research report MD file and just double click  on it. And you'll see it opens up. We have the uh chat over here with it. But here we have the  file. Now, at the top, I've got its objective or purpose here. A clear statement of what the  workflow does. Given a topic from the user, ask clarifying questions. Conduct structured research.  Okay. So, let's move down to the next. We have the qualifying question steps. Ask the user these  questions. And you get an idea. What are the specific angles of the questions you want covered?  Who will read the report? Do you want a quick uh overview or deep dive? And you can see the other  ones that include uh format preferences. Now, this is what makes it a gentic rather than just  a dumb text generator. It's going to check with you before it runs off and does the work. So,  no more garbage output because it guessed wrong about what you wanted. So, then it moves down.  We can see it confirmed the plan after receiving the answers. Confirm the research plan with users  before starting. We have step three, the research topic. as we uh go through and look at the details  for each of these, you know, like avoid opinion only posts. Uh for recent developments, include  sources from the last 12 months where possible uh what to look for while researching. Now we're  on to writing the report with the writing rules, the tone, and then saving the output. And we do  have some error handling. If the topic is too broad to research meaningfully, say so and ask the  user to narrow it. Now, here's the be beautiful part about all this. Every single word in this  file is just plain English. There's no code, no special syntax. If you wanted to change  something like adding a step where it creates a summary table or removing the clarifying questions  for quick task, you just edit the text. The agent adapts to whatever you write. So, that's the  power of this approach. I want to go back over to Claude Code and ask this question right here.  What workflows do you have available? And look, it lists our research report workflow and  describes exactly what it does. It knows what it's capable of now. So, let's put this to work.  Now, for the fun part, I'm going to give it a real ## Running the research workflow live [18:03] research task and we're going to watch the whole  agentic workflow run from start to finish. I'm going to go and put this in. I want to research  the current state of AI agents in 2026. What are people actually using them for? What's working?  What's overhyped? And where things are heading. Now, watch. The first thing it should do is ask  me questions because that's what the workflow says to do. So there it is. Look, it's asking,  you know, the audience, the scope, the depth, all these questions. Now, I need to answer it,  but I can answer it naturally, like I'm talking to a colleague, and this is how I'm going to do  it. Keep it broad. I want a hype versus reality breakdown. The audience is techsavvy people  who are curious about AI, but not necessarily developers. Medium depth. include specific tools,  platforms, stick to the last 6 to 12 months, and formatted structured posts for clear sections and  bulleted points. So, let's submit. Now, it's going to take those answers and start working. Watch the  to-do list. Now, here's the plan before it starts. You can see AI agents 2026, what people are  actually doing, all the uh things I just answered, what questions I'll research are these ones,  and I'm just going to say proceed. Yes. And notice now even on the side we have resources a  folder that's there. I'm going to say allow. Yes. So it's searching the web, reading through  results, pulling out relevant information. You can see it thinking through what it finds.  And remember this is web search built right into cloud code. No plugins, no MCP setup, no  BI API keys. It just works out of the box. Now I have to allow it to write. All right,  it's all done. The link to the finished report is here. It's also in my output. It lives right  here. It's the same thing. Now, the one thing I just wanted to point out here, this right here,  the uh most honest number, even the best models only succeed 45.7% of the time on real world task.  Human oversight isn't optional yet. I wonder how fast this number will change. I must say it's kind  of comforting to know this. But let's go ahead and open up the report. So remember when it opens  up the MD file, this is in markdown uh language here. So it doesn't look quite as nice as the  formatted report. But if we hit control, shift, and v together, it's going to give the preview of  it. So we have the topic and everything that we said said wanted in it from the depth of 3 to five  pages, the date range, the audience. We have our executive summary up top here, background,  uh what are the what's actually working, key findings. So, we have this detailed report now  going all the way through uh really narrowing down specifically what I told it to search. So,  each time I do this, it's going to be very specific. It gives me all the sources to this.  So, the agent did all the thinking, organizing, and all the writing. But we're not done yet. Let  me show you one more thing that makes the agendic ## Iterating and refining results [21:12] workflow so powerful. Iteration. I'm going to  say this. The executive summary is a bit long. Trim it to three bullet points. Now watch. It  doesn't start over again. It goes into the report, finds the executive summary, and trims it down.  Clean, targeted, edited. And one more thing I wanted to do, add a comparison table of the top  five AI agent tools mentioned in the report. So, look at that. It scans through what it already  wrote, identifies the tools it mentioned, and created a formatted comparison table. It didn't  need me to list the tools. It pulled it from its own research. Let's go check out those changes.  So, I'm just going to open back up this. I'll go back to preview. We should see a shorter summary.  There's the three. 1 2 3. And now, uh, before this was six. It didn't have quite the comparison  on it, but now 1 2 3 4 five on it. So, it made those quick updates to the files. That's what I  mean by Agent. It remembers the context and it knows the project and it builds on what's already  there instead of starting from scratch each time. Once you start working this way, it's really hard  to go back. All right, before we wrap up, let me ## Five mistakes to avoid and pro tips [22:22] save you some time with the five biggest mistakes  I see people make and some pro tips that go with them. Mistake number one, skipping claw.md. I  know I keep saying it, but this one file changes everything. Two minutes of setup saves you hours  of frustration. Just do it. Mistake number two, being too vague. Do some research gives you  generic garbage, but research remote working trends for general audience. Focus on productivity  data from the last 6 months and formatted as a structured report with bullet points. That gives  you gold. Being specific is everything. Mistake number three, not using plan mode. If you skip  planning and jump right to execution, the agent might go down the wrong path and you've wasted  your time. Plan first, build second. It takes an extra 30 seconds, but it's well worth it. Mistake  number four, not telling it to ask questions. If your workflow doesn't say ask clarifying  questions before starting, the agent will assume, and assumptions lead to wasted time and mediocre  output. Always build that check-in step. Mistake number five, trying to build everything at once.  Start with one workflow, get it working well, refine it, then build the next one. Don't try to  automate your entire life in a single weekend. Trust me on this one. Now, here are some quick  pro tips. Keep your workflow files in a workflow folder. It keeps things organized as you build  more of them. Read the agents to-do list and reasoning as it works. It's the best tool you have  for understanding what's happening and catching mistakes early before they snowball. When the  output isn't right, tell the agent specifically what to fix. Don't start over. Targeted feedback  is way faster than redoing the whole thing. And save your best claw.md and workflow files. You can  copy them into new projects as starting points. Over time, you'll build a library of workflows  that make it incredibly efficient. One more, and this is a really useful one. If the  agent goes off in the wrong direction, hover over a previous point in the thread, and  you have the option to roll back the conversation, and this undoes file changes that the agent made.  So, you can rephrase your prompt and try again. No need to manually undo edits or start over again.  It's a great safety net, especially when you're still learning how to write good prompts. And  finally, effort levels. Type effort followed by low, medium, or high to control how deeply  the agent thinks before responding. Low is fast and lightweight, great for simple edits. High  means it really slows down reasons carefully and produces its best work. Use that for more complex  research or multi-step task. If you're not sure, just leave it on the default and Claude will  figure it out. But knowing this dial exists is useful once you start running longer workflows.  Let's bring it all together. In the first video, ## Wrap up and weekly challenge [24:58] we used Claude code to build an app. Today, we  used it to build an agent. The first one does what you tell it. The second one figures things  out on its own. That's a huge jump. And that's what you did today. So, here's my challenge  for you. Build one workflow this week, just one. Pick something you do regularly, research,  writing, organizing, analysis, whatever it is, and turn it into a workflow file. It doesn't have  to be perfect. Just get it running and see what happens. Let me know down below in the comments  what you come up with. I'm always curious to see what people are thinking about, what they could  automate through one of these workflows. Thanks for watching this time on Teachers Tech. I'll see  you next week with more tech tips and tutorials.