MCP Debugger Tools
A VSCode extension that enables AI agents to programmatically control VSCode's debugging features through the Model Context Protocol (MCP).
README
MCP Debugger Tools
VSCode extension that exposes debugging capabilities via Model Context Protocol (MCP), enabling AI agents to programmatically control VSCode's debugger.
Features
The extension provides the following 4 MCP tools:
- add_breakpoint - Add a breakpoint to a specific file and line number
- remove_breakpoint - Remove a breakpoint from a specific file and line number
- start_debugging - Start debugging with a specific configuration
- stop_debugging - Stop the current debugging session
Installation & Setup
-
Clone and build the extension:
git clone <repository-url> cd vscode-mcp-debugger npm install npm run compile -
Install the extension in VSCode:
- Press
F5to run in Extension Development Host, or - Package as VSIX:
vsce packageand install manually
- Press
-
Verify the MCP server is available:
- Open VSCode Command Palette (
Cmd+Shift+P) - Run "MCP: List Servers"
- You should see "MCP Debugger Tools" listed
- Open VSCode Command Palette (
-
Enable in AI Chat:
- Open AI Chat in VSCode
- Click the tools icon or type
/to see available tools - You should see the 4 debugging tools available
Usage with AI Agents
Once installed, AI agents can use these tools to:
- Set breakpoints in your code at specific lines
- Start debugging sessions using your launch configurations
- Remove breakpoints when no longer needed
- Stop debugging to clean up resources
Communication Architecture
The extension uses a unique architecture to bridge the MCP protocol with VSCode's debugging APIs:
AI Agent → MCP Server → Temp File IPC → VSCode Extension → VSCode Debug API
↑ ↓
JSON Response ← Response File ← Debug API Result
How it works:
- AI Agent calls MCP tools (add_breakpoint, start_debugging, etc.)
- MCP Server (standalone Node.js process) receives the tool calls
- File-based IPC - Server writes command requests to temporary files
- VSCode Extension monitors temp directory and processes commands
- VSCode Debug API - Extension uses native VSCode debugging capabilities
- Response cycle - Results flow back through the same IPC mechanism
This architecture ensures the MCP server runs independently while having full access to VSCode's debugging features.
Example AI Agent Workflow
AI: I'll help you debug the issue in your calculator function.
1. First, let me add a breakpoint at the start of the function:
[Uses add_breakpoint tool with filePath and lineNumber]
2. Now I'll start debugging with your launch configuration:
[Uses start_debugging tool]
3. After analyzing the execution, I'll clean up:
[Uses remove_breakpoint and stop_debugging tools]
MCP Tools Reference
add_breakpoint
Description: Add a breakpoint to a specific file and line number
Parameters:
filePath(string, required): Absolute path to the filelineNumber(number, required): Line number (1-based)condition(string, optional): Conditional expression for the breakpoint
Example:
{
"filePath": "/path/to/your/file.js",
"lineNumber": 25,
"condition": "x > 10"
}
remove_breakpoint
Description: Remove a breakpoint from a specific file and line number
Parameters:
filePath(string, required): Absolute path to the filelineNumber(number, required): Line number (1-based)
Example:
{
"filePath": "/path/to/your/file.js",
"lineNumber": 25
}
start_debugging
Description: Start debugging with a specific configuration
Parameters:
configurationName(string, optional): Name of the debug configuration to useworkspaceFolderPath(string, optional): Path to the workspace folder
Example:
{
"configurationName": "Launch Node.js Program",
"workspaceFolderPath": "/path/to/workspace"
}
stop_debugging
Description: Stop the current debugging session
Parameters:
sessionId(string, optional): Optional session ID to stop a specific session
Example:
{
"sessionId": "debug-session-123"
}
Requirements
- VSCode with MCP support (latest version recommended)
- Node.js (v18+ for running the MCP server)
- Project workspace with debug configurations in
.vscode/launch.json - Extension Development Host (for development/testing)
Supported Platforms
- ✅ macOS - Fully tested and supported
- ✅ Linux - Compatible with file-based IPC
- ✅ Windows - Should work with standard temp directory
- ⚠️ WSL - May require additional configuration for file permissions
Debug Configuration Setup
Create a .vscode/launch.json file in your project:
{
"version": "0.2.0",
"configurations": [
{
"name": "Launch Node.js Program",
"type": "node",
"request": "launch",
"program": "${workspaceFolder}/src/index.js",
"console": "integratedTerminal"
}
]
}
Development
Building from Source
- Install dependencies:
npm install - Compile TypeScript:
npm run compile - Watch for changes:
npm run watchfor automatic compilation - Testing: Use the Extension Development Host (F5)
- Debugging: Set breakpoints in the extension code itself
Project Structure
src/
├── extension.ts # Main extension entry point & MCP registration
├── debugBridge.ts # VSCode Debug API wrapper
├── mcpServer.ts # Standalone MCP server (stdio transport)
├── integratedMcpServer.ts # Alternative integrated server implementation
└── vscodeDebuggerMcp.ts # Legacy server implementation
Key Components
- Extension.ts: Registers MCP server definition, monitors IPC files
- DebugBridge.ts: Provides clean interface to VSCode debugging APIs
- mcpServer.ts: Standalone MCP server that communicates via file IPC
- IPC System: File-based communication between processes
Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Test with Extension Development Host
- Submit a pull request
Please ensure all tools work correctly and add appropriate error handling.
Troubleshooting
Common Issues
-
Tools not visible:
- Check that MCP server is running via "MCP: List Servers"
- Verify the extension is active in Extension Development Host
- Look for "MCP Debugger Tools" in the MCP servers list
-
Breakpoints not working:
- Ensure file paths are absolute and files exist
- Check that the file is currently open or can be opened in VSCode
- Verify the line number is valid (1-based indexing)
-
Debug config errors:
- Verify
.vscode/launch.jsonexists and is valid - Ensure the specified configuration name exists
- Check that the workspace folder is correctly detected
- Verify
-
IPC communication issues:
- Check VSCode Developer Console for error messages
- Verify temp directory permissions (the extension uses OS temp directory)
- Look for MCP server stderr output in "MCP: List Servers" → Show Output
Debug Mode
Enable detailed logging by checking the MCP server output:
- Open Command Palette (
Cmd+Shift+P) - Run "MCP: List Servers"
- Select "MCP Debugger Tools"
- Click "Show Output" to see server logs
Technical Details
- MCP Server Process: Runs as standalone Node.js process via stdio transport
- Extension Process: Runs within VSCode extension host
- IPC Mechanism: Temporary JSON files in OS temp directory
- Response Timeout: 10 seconds (configurable in source code)
- File Cleanup: Automatic cleanup of temporary IPC files
License
MIT
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.