For anyone who spends a significant amount of time tinkering with Apple Shortcuts, the announcement at this year’s WWDC felt like a glimpse into a much more efficient future. Apple’s plan to integrate generative AI directly into the Shortcuts app—allowing users to simply describe their automation needs in plain English and have the system build it from scratch—is a massive leap forward. While the current process of dragging and dropping individual action blocks is undeniably powerful, it can also be an incredibly tedious and time-consuming endeavor. The prospect of using AI to handle the heavy lifting of construction is a game-changer for enthusiasts, offering a rapid starting point that we can then manually refine to perfection.
Of course, the reality of generative AI is that it is rarely perfect on the first try. However, for those of us who build complex workflows, having an automated “rough draft” to work from is infinitely better than starting with a blank canvas and clicking through endless menus. The primary drawback to this exciting vision is the wait; these features aren’t scheduled to arrive until this autumn with the rollout of the latest operating system updates. For power users and automation junkies, waiting several months for a feature that feels like a natural evolution of the platform is an exercise in patience—patience that some, thankfully, aren’t willing to suffer through.
Enter Federico Viticci, the founder of the essential Apple-centric blog MacStories. His enthusiasm for automation is legendary, and like many of us, he decided not to wait for Apple to catch up. Viticci took the initiative to build his own iteration of this technology, which he dubbed “Shortcuts Playground.” By leveraging powerful models like Claude Code or OpenAI’s Codex, he essentially created a DIY version of Apple’s upcoming feature. It is a brilliant example of the “do-it-yourself” spirit that has always defined the Apple enthusiast community, proving that if a tool doesn’t exist yet, you can often bridge the gap with a little ingenuity and the right technology.
Getting started with Shortcuts Playground requires a bit of technical comfort, specifically involving the Terminal. You’ll need to head over to the project’s GitHub repository to find the setup instructions, which involve copying and pasting a specific command line. Once you’ve crossed that threshold, the experience becomes incredibly intuitive. I’ve personally tested this through Claude Code—though it functions similarly via OpenAI’s Codex—and the process feels remarkably fluid. By simply typing a “/” followed by “shortcuts,” you are presented with a clean menu of options that effectively turn your command line into a shortcut-building architect.
The core functionality comes down to two primary modes: “build” and “remix.” If you are starting from a blank slate, the “shortcuts-playground:build” option is where the magic happens. You simply provide a conversational, plain-language description of what you want your automation to accomplish, and the agent begins mapping out the logic. It isn’t just a blind automation process, either; the agent acts like a collaborator, periodically pausing to ask clarifying questions or gently reminding you of the architectural limitations inherent within the Apple Shortcuts ecosystem. This back-and-forth ensures that, even if the result isn’t flawless, it is rooted in technical reality.
To put this to the test, I tasked the tool with a common daily workflow: compiling my local weather forecast, my current calendar schedule, and my active to-do list, all to be read aloud as a morning briefing. Watching the agent work through these layers was impressive; it quickly mapped out the necessary actions, translated my request into logical steps, and presented a functional build. While the convenience of official, system-integrated AI is still on the horizon, tools like Shortcuts Playground provide a fascinating, functional bridge. It reminds us that whether we rely on Apple’s upcoming official updates or the ingenious solutions created by the community, the future of personal automation is becoming more human-centric, more conversational, and a whole lot easier to navigate.
