VSC‑MCP: Bridging Language Server Protocol with Model Context Protocol

VSC‑MCP turns the power of Visual Studio Code’s Language Server Protocol (LSP) into a set of MCP tools that AI assistants can invoke over the network. By exposing LSP functionality as standardized tool calls, it lets an AI client read, analyze, and edit code in any language supported by VS Code without needing a local IDE. This solves the common developer pain point of having to run separate language servers or install heavy tooling for each project—everything is handled behind a single MCP interface.

The server offers two operational modes. In Standalone Mode it launches a lightweight TypeScript language server, ideal for quick JavaScript/TypeScript projects. In Docker Mode it runs a headless OpenVSCode Server inside Docker, exposing both the web UI on port 3000 and an LSP bridge on port 5007. The Docker image bundles a rich set of VS Code extensions (rust‑analyzer, pyright, eslint, etc.), giving the AI access to advanced language features such as type checking, linting, and refactoring across languages. The VSC‑MCP extension inside the server creates a WebSocket channel that forwards MCP tool calls directly to VS Code’s internal LSPs, while file‑I/O tools operate on the container’s filesystem.

For developers using AI assistants like Claude, this means a single MCP server can act as a code oracle for an entire repository. The assistant can request diagnostics, format files, or generate new modules, and the server will return results as if it were a fully‑featured IDE. Real‑time editing is possible by opening the web UI, allowing developers to see AI‑generated changes live. The server also enforces directory restrictions via the environment variable, ensuring that AI actions remain scoped to safe areas of the codebase.

Key capabilities include:

• Cross‑language support: any language with a VS Code extension can be queried through MCP.
• Unified LSP: eliminates the need to manage multiple language servers manually.
• Extension ecosystem: inherits VS Code’s vast plugin library for advanced tooling.
• Live collaboration: web UI enables real‑time inspection of AI edits.

Typical use cases are automated code reviews, incremental refactoring guided by an AI, or generating boilerplate for new features. By integrating VSC‑MCP into an AI workflow, developers gain a powerful, language‑agnostic code manipulation layer that scales from small scripts to large monorepos.