MCP server · Developer tools

AgentSensor MCP server

Lets your AI coding agent see the real crashes from your dev server and browser, mapped to the exact file and line.

Developer tools Coding For it and development Developer Tools

AgentSensor is a small helper that watches your development server and your browser for errors, then hands those errors to your AI coding assistant. Instead of copying and pasting error messages into a chat window, your assistant can look them up itself. It is handy for developers who use AI coding tools like Claude Code, Cursor, or Codex and want them to fix real bugs instead of guessing.

What is an MCP server? The 30-second version

On its own, your AI assistant can only chat with you. An MCP server is a small helper program that gives your AI a new skill or a connection to another tool. This one connects your AI to your running development server and browser, so it can see the actual errors happening there. When you ask it to fix a bug, it can look up the real crash details instead of asking you to paste them.

What this MCP server does

You start your normal development server through AgentSensor by putting a short command in front of it. AgentSensor then quietly watches the terminal output and the browser for any crashes or errors. When something breaks, it saves the details in a small local file, including the exact file and line where it happened. Your AI assistant can then ask AgentSensor for those errors and read the full details. After the AI makes a fix, it can check whether the error is gone.

What you can do with it

See a list of errors that are currently open in your dev session
Get the full stack trace for a specific error, with your own files highlighted
Read the last 30 log lines before a crash
Check whether an error came back after a code edit
Mark an error as resolved once it is fixed
Catch browser errors like React component crashes and unhandled promise rejections
Keep a status board so old fixed errors do not confuse your AI

Try asking your AI

“The checkout page just crashed, fix it.”
“Check runtime errors and fix the bug in examples/broken-app/server.ts.”
“Did the error I just fixed come back?”
“Show me the open errors from the last 10 minutes.”

What it gives back to you

Your AI gets back a list of open errors, each with a short ID, the file and line where it happened, how many times it occurred, and when. If it asks for details on one error, it gets the full stack trace plus the log lines leading up to the crash. After a fix, it gets a simple answer like still_failing or not_seen. The status board file is also updated so resolved errors stay resolved unless they happen again.

Before you start

What you need

Node.js version 18.17 or newer
An AI coding agent that supports MCP, like Claude Code, Cursor, or Codex
A Node.js, JavaScript, or TypeScript project with a dev server

Good to know

It only works with Node.js, JavaScript, and TypeScript projects for now, and it is still an early pre-release version, so expect things to change.

Install it with your AI

Add AgentSensor MCP server to your AI, no technical skills needed

You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.

Sign in to get the install prompt

Members get a ready-made prompt that lets the Claude desktop app check AgentSensor MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,609 MCP servers here for you.