# ServiceNow MCP Learning Project

This project is a small learning-focused example of a Model Context Protocol (MCP) server that connects to ServiceNow.

## Purpose

This repository is intended for:

- learning how MCP servers are structured
- experimenting with ServiceNow API interactions
- exploring how tools and context can be exposed to AI clients
- prototyping ideas in a simple, approachable setup

## What it does

The project demonstrates how to create an MCP server that can interact with ServiceNow records and expose capabilities to clients such as AI tools or local MCP inspectors.

## Prerequisites

- Node.js
- npm
- A ServiceNow instance
- A valid `.env` file with the required ServiceNow connection values

## Setup

1. Install dependencies:

   ```bash
   npm install
   ```

2. Create a `.env` file in the project root with the required ServiceNow configuration values.

3. Start the project in one of the supported modes:

   Stdio mode:

   ```bash
   npm run inspector
   ```

   HTTP mode:

   ```bash
   npm run start:http
   ```

   If you want to inspect the HTTP server with the MCP Inspector, run it in a separate terminal:

   ```bash
   npx @modelcontextprotocol/inspector
   ```

## Docker 

1. Build image
   ```bash
   docker build -t servicenow-mcp:latest .
   ```

2. Start the container
   ```bash
   docker run -d --name servicenow-mcp -p 3000:3000 --env-file .env servicenow-mcp:latest
   ```


## Notes

- This repository is meant to support learning and experimentation.
- Security, validation, error handling, and production hardening should be added before using this pattern in real environments.
- Use it as a reference or starting point rather than a fully hardened production implementation.