Google introduces enhanced capabilities for Gemini Code Assist: Universal access to Agent mode
In a significant move towards enhancing productivity for developers, Google has rolled out an update to the Gemini Code Assist Agent Mode in both IntelliJ and Visual Studio Code (VS Code). This update, announced in 2025, brings several enhancements to the table, including multi-step task completion, new control features, and improved integration with Integrated Development Environments (IDEs).
One of the key features in IntelliJ is the ability to pause running chat responses and the introduction of Thinking Tokens. These additions provide developers with more control over chat interactions and offer insights into Gemini's thought process.
The revised chat function in Code Assist, based on the Gemini-CLI, includes full project context, multi-file editing, and integrated tools. This means developers can now work on larger projects more efficiently, with the assistance of the AI agent.
The Gemini CLI works asynchronously in the background, triggered by events like new issues or pull requests. It has been updated and better integrated with VS Code, recognizing the context of opened files and accessing selected text.
In both IntelliJ and VS Code, the Agent Mode allows for a collaborative approach that combines the power of AI with the user's expertise. It addresses three key use cases: smart issue triage, accelerated pull request reviews, and on-demand collaboration. With native in-editor diffing functions, users can view proposed changes side by side and edit them before accepting.
The Agent Mode supports efficiency and code quality without taking the development process out of the hands of experts. It offers a collaborative approach that allows users to execute complex, multi-step tasks and goals with full control over every change.
In IntelliJ, the Agent Mode also allows the use of automatic approval mode, enabling the agent to act on the user's behalf. This feature is now available in the stable version of IntelliJ.
For those who are new to Gemini Code Assist, it's good news that users can try it for free. The release notes provide detailed information about the latest updates. A public roadmap for Gemini CLI v1 has been published to keep users informed about future developments.
The Gemini CLI GitHub Actions is now available and widely available in beta. Several improvements have been made in VS Code, including direct editing of code changes, improved clarity on code changes, persistent Agent Mode in chat history, optimized quota updates, real-time output of shell commands, support for approval and editing of tool calls in batch mode, faster UI performance and rendering, and Agent Mode status persisting between IDE restarts.
Lastly, the Agent Mode was initially available in the Insiders channel of VS Code and is now being provided in a public preview. The latest improvements to code completion in VS Code also allow for faster and more efficient programming. Additionally, the Agent Mode now supports custom slash commands, allowing for reusable input prompts to optimize interaction with the CLI.
With these updates, Gemini Code Assist continues to demonstrate its commitment to empowering developers with advanced AI-assisted tools, making complex programming tasks more manageable and efficient.
Read also:
- visionary women of WearCheck spearheading technological advancements and catalyzing transformations
- A continuous command instructing an entity to halts all actions, repeated numerous times.
- Oxidative Stress in Sperm Abnormalities: Impact of Reactive Oxygen Species (ROS) on Sperm Harm
- Is it possible to receive the hepatitis B vaccine more than once?
