Everyone Talking about MCP servers: Lets Understand

Introduction

Imagine a world where AI assistants can seamlessly access thousands of different capabilities—booking restaurants, searching maps, generating images, and maintaining memories across conversations—all without you having to switch between different applications. This world is rapidly becoming reality thanks to the Model Context Protocol (MCP).

In this post, I'll break down what MCP is, why it's causing such excitement in the AI community, and how it's fundamentally changing the AI landscape.

What is MCP?

Model Context Protocol (MCP) is an open standard released by Anthropic in November 2024 that allows AI models to communicate with external tools, services, and data sources. Think of it as a universal connector that lets AI assistants extend their capabilities far beyond what they can do on their own.

The Core Components

MCP has two main components:

1. MCP Hosts: These are applications that users interact with directly, like Claude Desktop, Claude Code, Cursor, and oterm terminals. They act as the interface between users and AI capabilities.

2. MCP Servers: These are the extensions that provide additional functionalities to the AI. There are now thousands of these servers, ranging from Google Maps integration to memory storage to image generation.

The brilliant part of MCP is that it's an open standard. This means any MCP host can work with any MCP server, creating an incredibly flexible ecosystem.

Why MCP Matters: Beyond Traditional Apps

MCP represents something fundamentally different from traditional app ecosystems like iOS or Android. Here's why:

1. Open Standard Extensibility

Unlike the iOS vs. Android divide, MCP creates a universal standard. A developer builds an MCP server once, and it works with any MCP host. This is why there are already thousands of MCP servers just months after the protocol's introduction. Even OpenAI, Anthropic's main competitor, adopted MCP in March 2025.

2. Seamless Integration and Chaining

Traditional apps are isolating—you open one app at a time. MCP enables AI hosts to:

• Chain multiple servers together
• Combine results from different servers
• Use the output of one server as input for another

This creates powerful workflows that were previously impossible or required complex custom development.

3. From Tools to Ecosystem

MCP represents a transition from isolated tools to an interconnected ecosystem. The most exciting part is that many components can act as both hosts and servers. For example, Claude Code can use GitHub as a server to check in code, but can also be used as a server by Claude Desktop to solve coding problems.

This points toward a future where we have a mesh of AI agents that can request services from each other and provide services to each other.

MCP in Action: A Real-World Example

Let me share a simple but powerful example of how I've used MCP:

I created a personalized news service using only natural language and MCP:

1. I told Claude Desktop about my interests
2. Claude stored this in the Memory MCP server
3. Claude uses web search to find relevant news
4. Claude remembers what it already told me to avoid duplicates
5. Each day, I get a custom news bulletin based on my interests

Without MCP, this would have required custom code, databases, APIs, and a web application. With MCP, I built it using just conversation.

How MCP Will Evolve

While MCP has made impressive progress, there are still challenges to overcome: