Visual Studio Code now supports Model Context Protocol (MCP) Apps, an extension that lets AI agents deliver interactive UI components directly inside the editor’s chat. The feature appears in the VS Code Insiders channel and is slated for a stable release within days, enabling developers to manipulate tables, charts, forms, and wizards without leaving the code window.
What Are MCP Apps?
MCP is an open protocol that standardizes how large language models interact with external tools and data. MCP Apps extend this protocol by allowing a tool call to include a structured UI description. VS Code fetches the UI resource, renders it in a sandboxed iframe, and enables bidirectional communication via JSON‑RPC, turning plain‑text responses into rich, interactive components.
How MCP Apps Work in VS Code
Built on the existing integration, MCP Apps let a tool return a UI payload such as:
{
“name”: “visualize_data”,
“description”: “Visualize data as an interactive chart”,
“_meta”: {
“ui”: {
“resourceUri”: “ui://charts/interactive”
}
}
}
VS Code retrieves the ui:// resource, displays it next to the chat message, and forwards user interactions back to the model. The sandboxed environment ensures security while preserving a seamless developer experience.
Demo Scenarios
- Interactive list reordering – Drag items in a list UI generated by the agent instead of typing commands.
- Configurable deployment wizard – Real‑time form adjustments let developers fine‑tune deployment parameters within the chat.
- Live data dashboards – Streaming charts update automatically as underlying data changes, eliminating the need to re‑run queries.
Benefits for Developers
By introducing UI‑driven interactions, MCP Apps reduce the “prompt‑response‑prompt” loop that dominates AI‑assisted coding. Developers can manipulate suggestions, configure environments, and explore data with familiar mouse and keyboard actions, cutting the time spent switching between VS Code, external dashboards, or terminals.
Third‑party vendors can create “AI‑first” extensions that leverage the MCP Apps standard, expanding the marketplace of AI‑powered utilities while maintaining security through sandboxed iframes and controlled JSON‑RPC channels.
Future Outlook
The upcoming stable release of MCP Apps signals a shift toward visual interactivity as a baseline capability for AI coding assistants. As more extensions adopt the protocol, developers can expect a unified experience where conversational AI and full‑featured applications converge inside VS Code.
