Switching from Studio to VS Code: Plugins, Debugging, and AI

One the main differences is the plugin system. Because in VS Code, everything is built around extensions. For example, for InterSystems development, you have the "InterSystems ObjectScript Extension Pack" that includes the essential tools. With this plugin you can connect to InterSystems servers, browse code, edit classes, compile, and work with ObjectScript directly from VS Code.

And it's not just for ObjectScript. VS Code also supports Git, Docker, REST clients, YAML, JSON, Python, and many other technologies. Instead of working in a closed IDE, you can build a workspace tailored to your actual project ! (https://code.visualstudio.com/docs/configure/settings)
Another major improvement is the debugging features. VS Code makes it visual and intuitive. You can set breakpoints directly in the editor. Which means you can pause execution, and step through the code to see what’s happening. This is especially useful when investigating business logic or tracking down unexpected behavior.
If you’re used to adding temporary logs, breakpoints might be a game-changer for you as you do not need to modify the code to understand what's going on. VS Code uses a standard debugging interface so if you were already using it with other languages you will have no problems using it with ObjectScript.
Then there is the AI integration (Not available for everyone yet). GitHub Copilot or Claude Code (for example) integrate directly into the IDE. They can generate boilerplate code, explain existing logic, suggest refactoring, write tests, or summarize unfamiliar code. For ObjectScript developers, this is handy when working with legacy code, documenting methods. It can also be connected to Jira or Confluence to help you update documentations or even help you understand what is requested in a Jira ticket. It can even be instructed with the InterSystems coding standards to ensure that the code you produce is compliant. (https://code.visualstudio.com/docs/agent-customization/custom-instructi…)
 

That said, AI should be seen as an assistant, not an authority. Generated code still needs to be reviewed, tested, and adapted to InterSystems conventions. The real value is not automatic coding, but better understanding, and more ease navigating the codebase.
When intagrated in VS Code, the AI has access to the codebase to add it in its context for better and more accurate answers, so you should be wary of the data you "feed" to the AI as to not let critical informations leak.
In the end, moving from Studio to VS Code is not about losing an old tool, it is about gaining a broader development environment. Plugins add flexibility, breakpoints make debugging more efficient, and AI opens up new ways to understand and improve your code. For InterSystems developers, VS Code is a practical step toward a more modern and collaborative workflow.