A Gentle Introduction to MCP
Or: What It Is, Why It Matters, and How You Can Actually Use It
Since its introduction in November 2024, something significant has been reshaping how we work with AI tools. Not through dramatic model upgrades or shiny new apps — but through something more subtle and powerful: MCP, the Model Context Protocol. It has since become mainstream, with major development tools and AI platforms adopting it at an accelerating pace.
If the term sounds technical or mysterious, you’re not alone. Many readers of this blog have written to me saying the same thing:
“Elena, I hear people talk about MCP in Cursor or Antigravity,
but… what is it actually? And do I need it?”
So today, we’ll take it slowly.
This post explains:
- what MCP really is (in human terms),
- why it’s becoming important,
- how popular AI tools use it,
- and how you can bring MCP into your writing or coding workflow — even if you’re not a backend engineer.
My goal is to help you feel comfortable with the idea that MCP is not a scary protocol.
It’s more like a polite assistant who knows how to knock.
🌿 What Is the Model Context Protocol?
Picture an AI assistant as a clever helper sitting at your desk. It’s not that AI can’t act at all without MCP — plenty of assistants already ship with their own file, browser, or code-execution tools. What MCP gives you is a standard way to add more capabilities, instead of every tool needing its own bespoke integration.
The protocol defines three roles: the host is the app you talk to (Claude Desktop, Cursor), a server is a small process that exposes some capability, and a tool is the specific action that server offers — create_draft, say. The specification also covers resources (file-like data a server can hand over) and prompts (reusable templates), though tools are what most people mean when they say “MCP”.
Where do the safety guarantees actually come from? Not from the protocol itself. MCP defines how hosts and servers talk to each other — message formats, capability negotiation, authorisation for remote servers — but it does not itself enforce approvals or file permissions. Those depend on the host you use (Claude Desktop asks before running a tool; not every host does), your operating system’s own permissions, and the server code itself. A local server runs with whatever permissions the process running it has.
MCP, in one line: a standard way for a host to connect an AI assistant to servers that expose tools, resources, and prompts — with the actual safety boundary set by the host and the server, not by the protocol.
This standardised model is why so many systems have adopted MCP: Cursor, Antigravity, Claude Desktop, and a growing list of agent frameworks.
What changed in MCP in 2026, when you actually need it, and a worked Python example follow below.
🌿 MCP Ecosystem Updates in 2026
🔒 Subscribe to keep reading.
🌿 When You Need MCP: Use Cases for Developers and Writers
🔒 Subscribe to keep reading.
🔒 Subscribe to keep reading.
🌿 How to Get Started with MCP: Installation Steps
🔒 Subscribe to keep reading.
🌿 MCP Security: Do You Need API Tokens?
🔒 Subscribe to keep reading.
🍃 Python MCP Server Example: Building a Blog Manager Tool
🔒 Subscribe to keep reading.
🍃 Connecting the Server to Claude Desktop
🔒 Subscribe to keep reading.
MCP Safety Checklist
🔒 Subscribe to keep reading.
References
🔒 Subscribe to keep reading.
You've hit a Deep Dive tutorial.
I spend dozens of hours researching, coding, and breaking things to write these guides. This content is free, but reserved for my subscriber community. Drop your email below to unlock this guide (and all past/future deep dives):
Full content temporarily unavailable — refresh in a moment
Already a subscriber? Use the magic link from your last newsletter, or reset your password.
Log in to unlock
New subscribers get an inbox mail: Set a password to unlock articles. The form does not log you in — use the same email afterwards.
