Learn how to turn interactive explorations into reusable workflows that run on Windows, macOS, or Linux, and how cloud services like BrowserBase let you automate browser tasks remotely. The shift is moving from prompts for exploration to prompts that execute directly, building AI scripts that work autonomously using MCP tools.
Chapters
because that allows you to combine the idea of you can do the exploration interactively and then compile that into a workflow that you can run just as a command whenever you want. And of course, once it's a command, it becomes something you can put on schedule to run whenever you want. And that is something you can run on like your, you know, on a Windows machine or a Mac machine or even a Linux machine. The question is like, how can you have something that runs completely in the cloud? That's where things get a little bit more tricky because a lot of this automation stuff for like controlling browsers is something that can be done through the cloud. Like there's services like BrowserBase to do this. And then you start taking this level up, right? And going from like the prompt-based exploration, which is what Demetrius was talking about, to maybe prompt-based execution, you know, an AI-centric script of, you know, getting the AI to do the work agentically and use tools from MCP to get it done.