Purple AI

Model Context Protocol

May 07, 2025

The Model Context Protocol (MCP), introduced by Anthropic, is an open interoperability standard that allows developers to establish secure, two-way communication between AI-powered applications and external data systems.

The Model Context Protocol (MCP) defines a standardized approach for integrating applications with language models by enabling them to:

  • Provide rich contextual information to enhance model understanding
  • Expose external tools, data, and services that AI systems can interact with
  • Construct flexible, composable workflows that combine multiple capabilities

At its core, MCP uses JSON-RPC 2.0 to facilitate structured communication between three key components:

  • Hosts: These are AI-powered applications, such as chatbots or agents, that initiate and coordinate connections.
  • Clients: Functioning within the host environment, clients act as intermediaries that handle specific data or tool interactions.
  • Servers: External systems or services that expose context, datasets, or executable functions for the AI to access and utilize.

MCP Servers act as bridges or connectors that offer a consistent way to access different external systems, tools, and data sources.

A variety of MCP Servers have been made available as open-source projects on GitHub, providing developers with ready-to-use implementations for integrating external tools, systems, and data sources with AI applications. These servers act as standardized interfaces that expose functionalities and contextual information to language models through the Model Context Protocol (MCP).

You can explore the full list of available MCP Server implementations and their documentation here:
🔗 https://github.com/modelcontextprotocol/servers

Use Case Scenarios

  • MCP Client A (Chatbot): Needs to fetch the latest code commits from a repository. It communicates with the MCP Server for GitHub, which in turn interacts with the GitHub API to retrieve the necessary information.
  • MCP Client B (IDE Plugin): Aims to access design documents stored in Google Drive. It connects to the MCP Server for Google Drive, facilitating seamless retrieval and integration of documents into the development environment.
  • MCP Client C (AI Agent): Requires data analysis on customer information stored in a PostgreSQL database. It utilizes the MCP Server for PostgreSQL to execute queries and process the results accordingly.

+-----------------+        +--------------------+        +--------------------+
|   MCP Client A     |         |   MCP Client B     |                   |   MCP Client C     |
| (e.g., Chatbot)    |         | (e.g., IDE Plugin) |                   | (e.g., AI Agent)   |
+---------+-------+        +---------+----------+        +---------+--------+
          |                                                                |                                                   |
          |                                                                |                                                   |
          |                                                                |                                                   |
          v                                                               v                                                  v
+--------------------+        +--------------------+        +----------------+
|   MCP Server:               |        |   MCP Server:                 |        |   MCP Server:        |
|   GitHub API                 |        |   Google Drive                |        |  PostgreSQL DB |
+--------------------+        +--------------------+        +----------------+
          ^                                                               ^                                                   ^
          |                                                                |                                                    |
          |                                                                |                                                    |
          |                                                                |                                                    |
+--------------------+        +--------------------+        +----------------+
|   External System:      |        |   External System:       |        |External System: |
|   GitHub                          |        |   Google Drive               |        |   PostgreSQL        |
+--------------------+        +--------------------+        +----------------+