GitLab Duo IDE拡張機能ガイド — VS Code / Visual Studio / JetBrainsで「まず使える」ようになるA practical setup guide for GitLab Duo IDE extensions covering VS Code, Visual…

GitLab Duo is GitLab's family of AI-assisted development features, and getting it running inside a code editor is the step that turns a subscription into everyday value. This guide walks through installing and first-configuring the GitLab Duo extensions for the three most common environments developers use: Visual Studio Code, Visual Studio, and the JetBrains family of IDEs such as IntelliJ IDEA, PyCharm, and GoLand. The goal is a consistent path to a working setup rather than an exhaustive tour of every capability.

The reason this matters is practical. AI coding assistance has become a standard part of many toolchains, and the difference between a smooth start and a frustrating one usually comes down to authentication, permissions, and knowing which settings to check first. Because GitLab Duo is tied to a GitLab account and its associated plan, the initial configuration involves more than simply installing an extension from a marketplace.

Before installing anything, it is worth confirming the prerequisites. GitLab Duo features are gated by subscription tier and by administrator settings, so access to Code Suggestions and Duo Chat depends on the plan attached to your GitLab.com account or self-managed instance. In organizational setups, an administrator may need to enable Duo for the group or project, and network policies or proxies can affect whether the extension can reach GitLab's services. Checking these points early tends to prevent the most common "it installed but nothing happens" problems.

For VS Code, the workflow centers on the official GitLab Workflow extension, which is available through the Visual Studio Code Marketplace. After installation, the extension prompts you to authenticate to your GitLab instance, typically by supplying an instance URL and a personal access token with the appropriate scopes, or through an OAuth-style sign-in flow. Once connected, Code Suggestions appear inline as you type, and Duo Chat is reachable from a dedicated panel. Users can usually adjust behavior, such as enabling or disabling suggestions per language, in the extension's settings.

The Visual Studio path uses a separate extension built for Microsoft's full IDE, installed through the Visual Studio Marketplace or the in-product extension manager. The authentication concept mirrors VS Code: point the extension at your GitLab instance and authorize it. Because Visual Studio is oriented toward .NET and C++ workflows, this integration is aimed largely at developers in the Microsoft ecosystem who still host code and pipelines on GitLab.

For JetBrains IDEs, the GitLab plugin is available from the JetBrains Marketplace and installs through the Plugins section of the IDE settings. After a restart, you configure the connection and token, and Duo features surface through the editor and a chat tool window. The plugin is designed to work across the JetBrains lineup, so the same general steps apply whether you are in IntelliJ IDEA, WebStorm, or another variant.

Across all three environments, the recurring themes are the same: install from the official marketplace, connect to the correct GitLab instance, grant a token with sufficient scope, and verify that Duo is enabled on the account. If suggestions do not appear, the usual suspects are an unauthenticated session, an insufficient plan, a disabled feature flag, or a corporate network blocking outbound requests.

It helps to place GitLab Duo in the wider context of AI coding tools. It competes with offerings such as GitHub Copilot, Amazon Q Developer, and editor-native assistants, and the broader industry has moved toward embedding large language model features directly into development environments rather than treating them as separate apps. GitLab's positioning appears to emphasize keeping AI assistance close to the rest of its platform, including source control, CI/CD pipelines, and issue tracking, so that suggestions and chat can draw on project context.

For teams evaluating these extensions, the sensible approach is likely to start with a small pilot, confirm data-handling and privacy settings that satisfy internal policy, and then standardize the setup steps documented here. A shared, repeatable configuration reduces support overhead and makes it easier for new contributors to reach a "just works" state quickly, which is ultimately what this kind of setup guide aims to deliver.