What is MCP?

Model Context Protocol (MCP) is an open standard that lets AI systems talk to external tools and services through a unified interface. Think of it as a universal translator that allows AI assistants to invoke functions, grab data, and perform actions consistently across different platforms.

MCP Overview

The Cast of Characters

MCP Clients

These are your AI assistants that connect to MCP servers:

Claude Desktop, Cursor, VS Code, Windsurf, etc.
Basically any AI tool that speaks MCP

MCP Servers

Applications (like our Bitly server!) that expose tools and capabilities to AI agents

Tools

The specific functions your AI can call to get stuff done

Resources

Data sources that can be read by the AI (e.g., configuration files, documentation)

Prompts

Pre-configured prompt templates for common tasks

MCP Transport Types

MCP supports different ways for clients and servers to chat with each other:

HTTP/Streamable HTTP (What Bitly Uses)

Remote servers accessible over the internet
No local installation required
Multiple users can connect at once
Communication through HTTP requests
Works well for cloud-based applications

stdio (Standard Input/Output)

Local servers running on your machine
Requires local installation
Communication through stdin/stdout
One user per instance

SSE (Server-Sent Events)

Legacy transport for remote servers
One-way server-to-client communication
Being phased out in favor of HTTP