ConsoleSpy MCP Server

ConsoleSpy MCP servers enable AI models to interact with browser console logs, providing capabilities for real-time debugging, error monitoring, and application analysis.

April 25, 2025
MCP ServerDevelopment Tools & DevOpsConsoleSpy MCP Server
GitHub stars

Overview

The ConsoleSpy MCP Server captures browser console logs and makes them available to AI models through the Model Context Protocol (MCP). This allows for real-time debugging, error monitoring, and in-depth application analysis directly within AI-assisted development environments.

Created by:

Developed by mgsrevolver

Key Features

👁️

Real-time Log Capture

Captures browser console logs in real-time for immediate analysis.

🐛

Enhanced Debugging

Provides AI models with direct access to application runtime information for debugging.

📊

Application Analysis

Enables AI to analyze application behavior and identify potential issues.

🔌

Seamless Integration

Integrates with Cursor IDE through the Model Context Protocol.

Available Tools

Quick Reference

ToolPurposeCategory
get_logsRetrieve captured console logsRead
clear_logsClear all captured console logsWrite

Detailed Usage

get_logs

Retrieve all captured console logs from the ConsoleSpy server.

use_mcp_tool({
  server_name: "consolespy",
  tool_name: "get_logs",
  arguments: {}
});

Returns an array of log entries.

clear_logs

Clear all captured console logs from the ConsoleSpy server.

use_mcp_tool({
  server_name: "consolespy",
  tool_name: "clear_logs",
  arguments: {}
});

Returns a confirmation message.

Installation

{
  "mcpServers": {
    "consolespy": {
      "command": "npx",
      "args": [
        "-y",
        "supergateway",
        "--port",
        "8766",
        "--stdio",
        "node",
        "console-spy-mcp.js"
      ],
      "env": {
        "CONSOLE_SERVER_URL": "http://localhost:3333/mcp"
      }
    }
  }
}

When to Use This Server

ConsoleSpy is useful in the same situations you'd otherwise spend minutes copy-pasting browser console output into a chat window. Because it captures logs continuously, an AI assistant can act on them without you switching tabs:

  • Debugging loops — Reproduce a bug in the browser, then ask the assistant to read the captured logs and propose a fix. No manual copy-paste between browser and editor.
  • Runtime error triage — When an error appears intermittently, the server keeps the recent console history available so the assistant can correlate the failing action with the logged output.
  • Regression checks — Run a feature, clear logs with clear_logs, exercise the flow again, and have the assistant review what changed.

For production error monitoring, pair it with Sentry MCP; for full browser automation and assertions, see Playwright MCP.

Frequently Asked Questions

Sources