Big updates in the latest @code release: ✨ Extend agent mode (now available in VS Code Stable!) with MCP tools ✨ A unified chat experience ✨ Use your own API key to access more language models ...and so much more aka.ms/VSCodeRelease Let’s dive in 🧵
Agent mode is now available in VS Code Stable. Enable it by setting chat.agent.enabled ⚙️ In the coming weeks, we will be rolling out enablement by default to all users.
@code Almost everything that @cursor_ai does 💀
@code Gotta say, pretty impressed with this so far. (Not sure about the expected pricing structure though…. More expensive than Windsurf…?)
@code What is the point considering how long Claude 3.7 and 3.5 have in preview on VSCode Insiders to now bring these models to the stable version and yet you don’t allow them to process images??? Make it make sense please.
@code Next edit function does not work well with the vi extension.
@code amazing updates, curious what will come next -- 2.0 or 1.100 :)
I just tried this. And it seems Agent Mode doesn't know about the other files we have, until we externally specify it. Even if I specify, it makes wrong decisions. I attached a file for reference, and some line selection where i wanted modification. It changed the reference file instead ! There is a long way to go. I'll wait.
@code Give us 70b models like llama for chat completions it shouldn't be limited or should be much cheaper for those of us who only need that
@code We need event based agents in copilot. webhook equivalent for mcps so to speak.
@code how about about exiting screens that duplicate better how about that. how about if a duplicate console of your entire project pops on your screen on the other duplicate screen you have a easier exit out button schedule a live for the none friendly bs you got cooking in there
@code I will avoid this version. I don't need it yet. 🤣
@code Hope you are going to consider Gemini 2.5 model , didn't see it in the list