View all Tracks
Learn to Build AI Agents
An agent is not a chatbot. A chatbot answers. An agent plans, acts, observes the result, and adjusts until the job is done.
In this track you'll build one from the ground up. You'll start by understanding the loop that makes agents different, then turn your idea into an agent project spec with a real job, clear inputs, defined limits, and a measurable definition of success.
You'll get hands-on quickly. Build your first working agent inside the Claude app, explore Claude Cowork for agent-style workflows, then extend what your agent can reach through connectors and computer use, always with supervision and permissions in mind.
From there you'll go deeper into the parts that make an agent dependable. You'll create reusable skills that make behavior consistent, use the /goal command for complex multi-step work and know when /loop is the better choice, sharpen your agent's prompt through testing, and choose the right model for the task using OpenRouter to compare quality, speed, and cost.
Then you'll equip it properly: tools connected through CLIs, APIs, and MCPs, persistent memory that respects user privacy, and guardrails that keep a capable agent from having unchecked authority. You'll finish by coordinating multiple agents, comparing the major agent platforms, and optionally giving your agent a friendly interface.
Start building your first AI agent with this Track!
Ready to start learning?
Treehouse offers a 7 day free trial for new students. Get access to 1000s of hours of content. Learn to code, land your dream job.
Start Your Free Trial
-
32 minWorkshop
Understand AI Agents and Plan Your First One - AI Agents Day 15
Learn what makes an AI agent different from a simple chatbot or one-off prompt. You’ll explore the loop agents use to plan and act, then write a clear project specification that gives your first agent a useful direction and measurable outcome.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
17 minWorkshop
Do Agentic Work with Claude Cowork - AI Agents Day 16
See what Claude Cowork can do with a real multi-step task. You’ll give it a spreadsheet and a goal, watch it work through the steps, and review the files it produces, then set the project context and schedule so the work can run without you.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
12 minWorkshop
Extend Your Agent with Connectors and Computer Use - AI Agents Day 17
Move beyond an agent that only reads and writes text. Learn how connectors provide access to useful services and how computer use can let an agent interact with a screen on your behalf, while keeping control and safety in mind.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
16 minWorkshop
What AI Agent Skills Are and How to Create One - AI Agents Day 18
Learn how skills package repeatable instructions and expertise for an AI agent. You’ll understand when skills are useful and create your first one so an agent can perform a focused task more consistently.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
9 minWorkshop
Plan Complex Agent Tasks with the /goal Command - AI Agents Day 19
Learn how the /goal command helps an agent work toward a larger outcome. Compare goal-based planning with the /loop command, and choose the right approach for short repeated work versus more complex multi-step tasks.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
30 minWorkshop
Define Your Agent’s Goal and Improve Its Prompt - AI Agents Day 20
Decide exactly what your agent should do before you build it. You’ll define useful outcomes, clarify boundaries, and revise your prompt so a custom agent has the context it needs to behave consistently.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
16 minWorkshop
Build a Custom AI Agent - AI Agents Day 21
Bring your plan and prompt together to create a custom AI agent. You’ll scaffold the project with Codex and Vercel’s Eve framework, define an orchestrator and its specialist subagents, and add a brand profile that gives every part of the system the same context.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
45 minWorkshop
Choose an AI Model for Your Agent with OpenRouter - AI Agents Day 22
Learn how model choice affects an agent’s quality, speed, cost, and capabilities. Compare the factors that matter for different tasks, explore OpenRouter as a way to reach many models, then connect a model to the agent you built.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
20 minWorkshop
Connect Your AI Agent to APIs and MCP Servers - AI Agents Day 23
Give your agent practical ways to interact with the world. Learn how CLIs, APIs, and MCPs connect an agent to services, then connect one yourself so you can reach your agent from your phone.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
17 minWorkshop
Give Your AI Agent Persistent Memory - AI Agents Day 24
Learn how an agent can retain useful information across interactions. Explore where memory can live, how to decide what gets saved, and how a real-time database keeps memories and history in sync across the terminal, messaging, and an interface you build later.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
35 minWorkshop
Add Guardrails and Safety to Your AI Agent - AI Agents Day 25
Design an agent that is useful without giving it unchecked authority. Learn how tight instructions, a minimal tool surface, adversarial testing, human approval, and a guard layer keep your agent from drifting out of scope or acting on hostile content.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
11 minWorkshop
Manage Multiple AI Agents and Compare Platforms - AI Agents Day 26
Explore what changes when one task becomes a team of agents. Learn how to divide responsibilities, coordinate work, and review results, then survey popular alternatives so you can choose tools that fit your goals.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
24 minWorkshop
Create a GUI for Your AI Agent - AI Agents Day 27
Give your agent a friendly interface that makes its capabilities easier to discover and use. This optional workshop covers the design decisions behind a simple GUI, including inputs, progress, results, errors, and user control.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
9 minWorkshop
Reflect on Your AI Agent Journey - AI Agents Day 28
Look back on the full agent-building journey and identify the ideas, tools, and safety practices that shaped your work. Reflect on how you now think with AI, then create a practical plan for moving forward as an AI-native agent builder.
This is part of the AI Bootcamp. Click here to learn more. (/ai-bootcamp)Viewed -
2 minInstruction
Congratulations — You've Completed the AI Bootcamp
You've completed the AI Bootcamp. Look back at what you built from a blank page, and at the AI-native habits you'll carry into every project after this one. Then find out where to go next: the two remaining build tracks, why the second one goes faster than the first, and the one small thing to build this week....(continue reading)
Viewed -
60 minExam
AI Bootcamp - AI Agents Track Final Exam
Put your AI Agent Building skills to the test. This final exam includes a mix of question types to assess what you've learned. Pass to earn your certificate.
-
Track Completion
This track includes:
- Understand AI Agents and Plan Your First One - AI Agents Day 15 32 min
- Do Agentic Work with Claude Cowork - AI Agents Day 16 17 min
- Extend Your Agent with Connectors and Computer Use - AI Agents Day 17 12 min
- What AI Agent Skills Are and How to Create One - AI Agents Day 18 16 min
- Plan Complex Agent Tasks with the /goal Command - AI Agents Day 19 9 min
- Define Your Agent’s Goal and Improve Its Prompt - AI Agents Day 20 30 min
- Build a Custom AI Agent - AI Agents Day 21 16 min
- Choose an AI Model for Your Agent with OpenRouter - AI Agents Day 22 45 min
- Connect Your AI Agent to APIs and MCP Servers - AI Agents Day 23 20 min
- Give Your AI Agent Persistent Memory - AI Agents Day 24 17 min
- Add Guardrails and Safety to Your AI Agent - AI Agents Day 25 35 min
- Manage Multiple AI Agents and Compare Platforms - AI Agents Day 26 11 min
- Create a GUI for Your AI Agent - AI Agents Day 27 24 min
- Reflect on Your AI Agent Journey - AI Agents Day 28 9 min
- Congratulations — You've Completed the AI Bootcamp 2 min
- AI Bootcamp - AI Agents Track Final Exam 60 min