How to continue a GitHub Copilot app session in VS Code

GitHub Copilot now lets you seamlessly move conversations from the GitHub Copilot app (available on web and mobile) directly into VS Code while preserving your entire session context. This means you can start planning an architecture on your phone during a commute, then pick up exactly where you left off in your IDE without restarting the conversation or losing the code suggestions you’ve already received. It’s a small feature that removes friction from how we actually work across devices.

Under the hood, this works through session linking between GitHub’s hosted Copilot service and your local VS Code environment. When you initiate a “continue in VS Code” action from the Copilot app, GitHub generates a session token that includes your conversation history, the code context you’ve been discussing, and any artifacts or code blocks generated so far. Your VS Code instance retrieves this context using the GitHub Copilot extension, which authenticates with your GitHub account and loads the session state. Technically, it’s similar to how cloud IDEs maintain state—the conversation isn’t being re-synced in real-time, but rather a point-in-time snapshot is transferred and loaded into your local environment.

The practical value here depends on your workflow. If you’re doing exploratory work—sketching out a Lambda function design, planning a database schema, or debugging infrastructure-as-code—you might start that conversation on the Copilot app while away from your desk. Once you’re back at your computer and ready to actually implement, continuing in VS Code means Copilot already understands what you’re building and can provide more targeted code completion and suggestions. You avoid the cognitive overhead of re-explaining the problem. For teams using Copilot in their deployment pipelines or for infrastructure automation, this continuity also reduces the chance of context-switching errors where you forget a requirement you’d discussed on mobile.

The feature is particularly useful for people juggling multiple projects or working across environments. A DevOps engineer might use the mobile app to draft runbooks during on-call rotations, then refine them in VS Code once back at a terminal. A Python developer debugging an API integration could ask questions on the web app, then immediately test solutions in their local environment without losing the thread of conversation. It’s a small but meaningful step toward AI assistance that follows you across devices without forcing you to restart from scratch.