Back to MCP Catalog

Claude Debugs For You MCP Server

Developer ToolsTypeScript
Interactive code debugging with LLMs through VS Code
Available Tools

startDebugSession

Start a debugging session for the current file or project

evaluateExpression

Evaluate an expression in the current debug context

getVariables

Get the variables in the current debug context

stepOver

Step over the current line of code

stepInto

Step into the function call at the current line

stepOut

Step out of the current function

continue

Continue execution until the next breakpoint

pause

Pause execution

Claude Debugs For You is a powerful debugging assistant that enables any LLM (like Claude) to interactively debug your code through VS Code. This MCP server and VS Code extension combination allows AI models to evaluate expressions, inspect variables, and help troubleshoot issues in real-time. The tool is language-agnostic, working with any programming language that supports debugger console functionality in VS Code. It creates a seamless debugging experience where your AI assistant can directly interact with your running code, making debugging sessions more efficient and insightful.

Overview

Claude Debugs For You (also known as "Vibe Debugging") creates a bridge between your LLM assistant and VS Code's debugging capabilities. This allows AI models like Claude to interactively debug your code, evaluate expressions, and help you solve problems more efficiently.

Installation

VS Code Extension Setup

  1. Install the VS Code extension from the VS Code Marketplace or download the .vsix file from the GitHub releases page.

  2. If installing from a .vsix file:

    • Open VS Code
    • Go to the Extensions view (click the Extensions icon in the Activity Bar)
    • Click the "..." menu in the top-right of the Extensions view
    • Select "Install from VSIX..." and choose the downloaded file
  3. After installation, you'll see a new status menu item "Claude Debugs For You" in your VS Code status bar, showing whether the service is running properly (check mark) or failed to start (x mark).

MCP Configuration

To use Claude Debugs For You with your LLM client, you need to configure the MCP server. The setup varies depending on your preferred connection method:

For Claude Desktop (using stdio):

Add the following configuration to your Claude Desktop settings:

{
  "mcpServers": {
    "claude-debugs-for-you": {
      "command": "code",
      "args": ["--stdio-mcp"]
    }
  }
}

For other LLM clients (using HTTP):

  1. Start the MCP server by clicking on the "Claude Debugs For You" status bar item and selecting "Start MCP Server".
  2. Note the URL displayed (typically http://localhost:8080).
  3. Configure your LLM client to use this URL as an MCP server.

Usage

  1. Open your project in VS Code.
  2. Ensure you have a valid launch.json configuration for debugging your project.
  3. Start a debugging session in VS Code.
  4. Connect to the MCP server from your LLM client (Claude Desktop, Continue, or other compatible clients).
  5. Ask your LLM to help debug your code. For example:
    • "Can you help me debug this function?"
    • "What's the value of this variable at this breakpoint?"
    • "Evaluate this expression in the current context"

The LLM will be able to:

  • Evaluate expressions in the debug console
  • Inspect variable values
  • Step through code
  • Analyze the execution state
  • Suggest fixes based on the runtime information

Troubleshooting

If you encounter issues:

  1. Check the "Claude Debugs For You" status in the VS Code status bar.
  2. Try restarting the MCP server from the status bar menu.
  3. Ensure your debugging session is active in VS Code before asking the LLM to debug.
  4. Verify your launch.json configuration is correct for your project.

For more detailed help, visit the GitHub repository.

Related MCPs

Apple Shortcuts
Developer ToolsJavaScript

Control Apple Shortcuts automations from AI assistants

Clojars Dependency Lookup
Developer ToolsJavaScript

Fetch dependency information from Clojars, the Clojure community's artifact repository

Simple Timeserver
Developer ToolsPython

Provides Claude with current time and timezone information

About Model Context Protocol

Model Context Protocol (MCP) allows AI models to access external tools and services, extending their capabilities beyond their training data.

Generate Cursor Documentation

Save time on coding by generating custom documentation and prompts for Cursor IDE.