{"id":33531,"date":"2024-02-06T19:42:28","date_gmt":"2024-02-06T19:42:28","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/cppblog\/?p=33531"},"modified":"2024-02-23T22:02:13","modified_gmt":"2024-02-23T22:02:13","slug":"quick-tips-for-using-copilot-chat-with-c-in-visual-studio","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/cppblog\/quick-tips-for-using-copilot-chat-with-c-in-visual-studio\/","title":{"rendered":"Quick Tips for using Copilot Chat with C++ in Visual Studio"},"content":{"rendered":"

GitHub Copilot Chat is now available in both Visual Studio<\/a> and Visual Studio Cod<\/a>e<\/a> when you have an active subscription to the service. Copilot Chat is your AI pair-programmer that allows you to speak in plain text through the chat interface to accomplish programming tasks such as understanding sections of code, generating test cases, and generating code.<\/p>\n

This blog post will focus on tips for using Copilot Chat functionality with C++ in Visual Studio, but the same functionality is available in Visual Studio Code and will be covered in future blog posts. Additionally, in future posts, we will dive into various C++ scenarios as reported by our customers.<\/p>\n

You can learn more about the Copilot Chat extension through the Visual Studio documentation<\/a>.<\/p>\n

\n

Download GitHub Copilot Chat<\/strong><\/h4>\n

To learn more about accessing GitHub Copilot Chat in Visual Studio, please view these instructions<\/a>. You will need a subscription<\/a> to GitHub Copilot for Individuals or GitHub Copilot for Business. You will also need both the Copilot<\/a> and Copilot Chat<\/a> extensions installed for autocomplete-style suggestions and the chat-based features respectively.<\/p>\n

\n

Introduction to Copilot Chat in Visual Studio<\/h3>\n

You can use the new Copilot Chat functionality to work with your existing C++ code. \u00a0To open the chat window, navigate to View > GitHub Copilot Chat <\/em>or right-click anywhere in your code and select Ask Copilot<\/strong>.<\/p>\n

\"An<\/a><\/p>\n

Performing actions via \/ commands<\/h5>\n

In the chat window, you can use the \/<\/strong> symbol to list out commands to perform on the codebase. The full list of available commands can be found here<\/a>, or the list of commands will automatically populate the available commands when you type \/<\/strong> in the chat window.\"A<\/a><\/p>\n

For example, you can use \/explain<\/strong> to get code explanations on a desired function or file. You can also use \/generate <\/strong>to ask Copilot Chat to generate a specific snippet of code.<\/p>\n

Adding file context via # sign<\/h5>\n

Large language models (LLMs) are trained on public repository data at a point in time and do not know anything about your current codebase. You can use the #<\/strong> commands to specify the context you want Copilot Chat to include and help Copilot Chat provide contextually relevant answers. For example, you can specify #file:FIleName.cpp<\/strong> in-line in your prompt to Copilot Chat so that it looks at the specified file. Once you type the\u00a0#\u00a0<\/strong>sign, all file names or selections will populate to select.<\/p>\n

\"Pressing<\/a><\/p>\n

By selecting a specific context, you can provide more information to Copilot Chat without having to copy\/paste code and ensure you receive the most relevant answers when applicable. Copilot Chat will always indicate which files were included as context under References<\/strong> under your prompt. For example, by default, there are no references sent from my code upon no file specification.<\/p>\n

\"Copilot<\/a><\/p>\n

Then, once some file is specified, you can see the code references in the References<\/strong> flyout.<\/p>\n

\"Image<\/a><\/p>\n

As you’ll see, Copilot Chat can empower you as a C++ developer to focus more on the task at hand by being a true AI pair programmer through help with asks such as:<\/p>\n