Amazon Q brings built-in AI-powered chat right into the code editor for developers

Amazon Q DeveloperArtificial Intelligence Software Development Assistant at Amazon Web Services Inc. announced today support for built-in chat, which combines the ability to call assistants in software editors and code updates, allowing developers to directly describe their needs and receive answers.

Many AI software development tools work as native code additions and as sidebar chat tools. Built-in code completion displays suggestions as developers type, allowing them to simply accept or reject them as they go, saving time. The chat interface helps generate code from developer prompts and also explains why, and also helps improve the code, including refactoring or creating tests and documentation.

Built-in chat allows developers to bring up a chat interface directly in their code editor and talk to the AI ​​assistant to tell it what they want to do. For example, they can tell him that they want to improve some segment of the code. directly on the spot. The advantage is that instead of need to use chat interface in the sidebar, they can immediately merge the proposal rather than copy/paste changes.

“This new capability is ideal for editing an existing file to fix problems, optimize code, refactor code, and add comments,” said Yose Yapir, Senior Developer Advocate at AWS. “Built-in chat Really powerful and helps me do more complex things quickly and accurately.”

Under the hood, Q Developer is built on Amazonthe company’s managed service for building AI applications and custom, high-performance base models, including from Amazon and other leading AI companies. To provide the best experience for developersAmazon said the built-in chat works on the latest version of Anthropic PBC. Claude 3.5 Sonnetwhich was launched a week ago.

The newest version of Sonnet is one of the most effective AI models in codesheets, showing significant improvements on industry benchmarks. This is the top model at the moment SWE diagramsolving 49% of the validated dataset, which is 500 real GitHub issues.

“By delivering a new built-in chat experience with this cutting-edge Anthropic model, Amazon Q Developer provides an AI assistant that can help you save time on your toughest coding challenges with unprecedented capabilities,” Yapir said.

Using the new built-in chat is as easy as can be and feels like engaging another developer in the process. Developers only need to select a segment of code and invoke Q Developer using ⌘ + I on Mac or Ctrl + I on Windows.

A small chat interface will appear in the editor where the developer can enter a prompt such as “Add documentation, including description, throughout the code.” The assistant will then analyze the code and create appropriate comments. in code.

Once complete, the user can review the AI ​​assistant’s performance and accept or reject the changes. In most cases, a block of comments is created at the beginning of the code documenting the software to explain how it works, and then inline elements are placed to explain anything that might require additional attention.

Although it is possible to create documentation using the chat interface, the user will have to copy and paste each comment block manually and with great difficulty. According to AWS, this makes the task easy and simple.

Built-in chat capabilities using artificial intelligence are available for the independent development environments Visual Studio Code of Microsoft Corporation and JetBrains sro. Developers can access the new feature through the free Amazon Q Developer tier.

Image: SiliconANGLE/Microsoft Designer, AWS.