Hands‑On AI Agents: Vibe‑Code AI Workflows and Agent Skills for Personal Productivity

Experience Level: Skill Building

Vibe-code your own personal agent. There's a lot of hype around AI agents that answer emails, do research, organize tasks, and maybe even run parts of your life. But how do these agents actually work? Can you trust them? What happens to your data? And where do they actually help versus quietly create new problems?

This Workshop cuts through the hype and helps you learn to build AI agents to truly scale your productivity. We'll explain what agents are, what works, what does not, what is safe, and what is not. We start by demystifying how LLMs operate so you can understand why they sometimes hallucinate, where “reasoning” breaks down, and why good results usually come from structure, not magic prompting. From there, we'll build a practical vocabulary for agentic systems: how they plan, how they use tools, how memory works, and why most consumer demos fall apart in real workflows.

In the second half, we'll get hands-on with Claude's consumer tools and work through real examples like email triage and LinkedIn tasks. In breakout groups, we'll turn repetitive work into structured, reusable AI workflows that last beyond a single chat session, without writing a line of code.

You do not need coding experience. To fully participate, you will need a Pro subscription from Claude, the Claude Chrome extension, and ideally the Claude desktop app. We will discuss other alternatives and free options but we'll build using those tools.