Google just made its AI coding agent, Jules, way more useful for developers

Jules, Google’s AI-powered coding agent, has been a powerful tool for developers, helping with everything from generating code and fixing bugs to writing tests and improving performance. But until now, using it meant working primarily through its dedicated website or a GitHub interface. In a step toward making Jules a true, seamless collaborator, Google is now bringing its AI agent directly into the developer’s natural habitat with a new command-line interface and a powerful API.

The first major update is Jules Tools, a new lightweight command-line interface (CLI). This allows developers to bring Jules directly into their terminal, where they can start, stop, and verify tasks right alongside their own commands. This is a game-changer for workflow efficiency. It’s the simplest way to move from chatting with Jules on a website to having it work alongside you in your actual development environment, right where the coding happens.

For even deeper integration, Google is also opening up access to the Jules API. This makes it possible to wire Jules directly into your own systems and workflows. The company gives examples like triggering tasks automatically when a bug is filed in Slack or integrating Jules into a CI/CD pipeline for automated testing and fixes. The API is all about putting Jules closer to the way development teams already build, making it a more organic part of the process. Support for Google Workspace users is also slated to arrive later this month.

These new access points are built on a foundation of recent improvements to the core experience. Over the past few weeks, Google has been focused on improving Jules’ reliability and quality, reducing latency, and fixing common setup issues. They’ve also rolled out key features like a File Selector to provide tighter context in chat, Memory so Jules can remember your preferences over time, and a structured way for Jules to access Environment Variables.