VSCode Debug MCP
An MCP server and VS Code extension that enables AI clients to interactively debug code using breakpoints, execution control, and state inspection. It is language-agnostic and works with any debugger that supports VS Code's launch.json configurations.
README
<img src="./images/vscode-debug-mcp-logo-128.png" width="64" height="64" alt="VSCode Debug MCP" align="center"> VSCode Debug MCP
Enable an MCP Client to interactively debug code with breakpoints, expression evaluation, and program launching.
This is an MCP server and VS Code extension that exposes debugging capabilities via the Model Context Protocol. It's language-agnostic — it works with any debugger that has a valid launch.json in VS Code.
Tools
| Tool | Actions | Description |
|---|---|---|
debug_execute |
launch, stop, continue, stepOver, stepIn, stepOut | Control program execution during debugging |
debug_breakpoints |
set, remove, list | Manage breakpoints with support for conditions, hit counts, and log messages |
debug_inspect |
evaluate, stackTrace | Inspect program state while paused at a breakpoint |
Getting Started
- Download the extension from releases
- Install the extension
- If using
.vsixdirectly, go to the three dots in "Extensions" in VS Code and choose "Install from VSIX..."
- If using
- Once installed, you'll see an "MCP Debug Server" panel in the Debug tab showing the server status, stdio path, and SSE address.

Claude Code
- Copy the ready-to-paste setup command using the "Copy Claude Code Setup Command" button in the Debug tab's MCP Debug Server panel, or via the VS Code command palette ("VSCode Debug MCP: Copy Claude Code Setup Command").
- Paste and run it in your terminal:
claude mcp add --transport stdio vscode-debug -- node /path/to/mcp-debug.js
Or add it to your project's .mcp.json:
{
"mcpServers": {
"vscode-debug": {
"command": "node",
"args": ["/path/to/mcp-debug.js"]
}
}
}
GitHub Copilot
- Copy the stdio server path to your clipboard using the VS Code command "Copy MCP Debug Server stdio path to clipboard"
- Open the VS Code command palette and run "MCP: Open Workspace Folder Configuration" to edit
.vscode/mcp.json, then add:
{
"servers": {
"vscode-debug": {
"type": "stdio",
"command": "node",
"args": ["/path/to/mcp-debug.js"]
}
}
}
Configuration
mcpDebug.port— Server port (default:4711)mcpDebug.autostart— Automatically start the server on activation (default:true)
When multiple VS Code windows are open, the extension handles graceful hand-off of the MCP server between windows.
Architecture
graph LR
subgraph MCP Clients
A[Claude Code / stdio client]
B[Cursor / SSE client]
end
subgraph Stdio Bridge
C[mcp-debug.js]
end
subgraph VS Code Extension
D[DebugServer]
E[VS Code Debug API]
end
F[Any Debugger<br/>Python, Node.js, C++, Go, ...]
A -- stdio --> C
C -- HTTP POST /tcp --> D
B -- SSE /sse --> D
D --> E
E --> F
The extension provides two ways for MCP clients to connect:
- Stdio — A standalone Node.js bridge process (
mcp-debug.js) receives MCP requests over stdio and forwards them via HTTP to the extension's debug server. Used by Claude Code and other stdio-based clients. - SSE — Clients connect directly to the extension's SSE endpoint. Used by Cursor and other SSE-capable clients.
Both paths use the VS Code Debug API under the hood, which means the extension works with any language/debugger that supports launch.json configurations (Python, Node.js, C++, Go, etc.).
Developing
- Clone and open this repo in VS Code
- Install dependencies:
npm install && cd mcp && npm install && cd .. - Build:
npm run compile - Press F5 to launch the extension in a new VS Code window
- Rebuild after changes:
npm run compile
Packaging
vsce package
Release Process
This project uses conventional commits and release-please for automated versioning and releases.
Commit Messages
All commits should follow the conventional commits format:
fix:— Bug fixes (patch version bump)feat:— New features (minor version bump)feat!:/fix!:/refactor!:— Breaking changes (major version bump)chore:,docs:,refactor:,test:— No version bump
PR Testing
Every pull request automatically builds a .vsix artifact via CI. Download it from the GitHub Actions "Artifacts" section on the PR's checks to install and test the extension locally before merging.
Releasing
When conventional commits are pushed to main, release-please automatically opens (or updates) a Release PR with a version bump and changelog entry. Merging that PR creates a GitHub release and builds the final .vsix, which is uploaded to the release assets.
Acknowledgements
This project is a fork of Claude Debugs For You by Jason McGhee. Thank you for creating the original extension and MCP server that this project builds upon.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.