Auto-Starting MCP Servers in Visual Studio Code Just Got a Whole Lot Easier
700The latest update to Microsoft’s popular code editor, Visual Studio Code (VS Code) version 1.103, brings with it several exciting features that make working on large projects even more efficient.
One of the most notable additions is the ability to configure auto-start behavior for Model Context Protocol (MCP) servers. This means users no longer have to manually restart MCP servers after updates, saving them time and effort in the long run.
Auto-Starting MCP Servers: A Game-Changer
For those who may not be familiar with MCP, it’s a standard protocol that allows applications to provide context to large language models. In simpler terms, it enables VS Code users to access and utilize various tools and features within the editor.
The newly added support for auto-starting MCP servers means developers can now focus on their code without having to constantly restart servers or worry about losing work in progress. This feature is especially useful when working with multiple projects or collaborating with team members, as it eliminates the need for manual server restarts and ensures seamless workflow.
In related news, Microsoft has also enabled an experimental tool-calling mode within VS Code. When dealing with a large number of tools (over 128), this feature automatically groups them together and allows developers to activate and call entire groups at once. This makes it easier for users to manage their tools and reduce clutter in the editor.
More Features, More Efficiency
VS Code version 1.103 also introduces support for Git worktrees, which enable checking out of multiple branches simultaneously. Worktrees allow developers to test changes or work on parallel projects without having to switch contexts constantly, making it easier to manage complex codebases.
The editor now detects and displays worktrees in the source control repositories view when opening a folder or workspace with a Git repository. Users can easily create, delete, open, or close worktrees using commands from the Command Palette or Source Control Repositories view.
What’s New for Copilot Chats?
Copilot chats in VS Code version 1.103 have also received some attention with the introduction of checkpoints that allow users to restore different states of chat conversations. This feature is particularly useful when working on multiple files during a session, as it enables developers to revert edits and go back to specific points in the conversation.
Additionally, GPT-5 (OpenAI’s newest model) has been rolled out for all paid GitHub Copilot plans, offering improved capabilities compared to its predecessor. The chat provider API has also undergone changes, allowing users to select which models appear in their model picker. Math rendering is now available in preview mode within chat responses.
Overall, the VS Code 1.103 update brings a range of exciting features that make working with large projects and language models more efficient than ever before.
What do you think?
It is nice to know your opinion. Leave a comment.