Skip to content

Core Concepts

This page explains the key concepts behind mcpconn and the protocol it is built upon.

The Model Context Protocol (MCP)

At its heart, mcpconn is designed to simplify interactions with servers that use the Model Context Protocol (MCP). MCP is a standardized protocol that creates a common ground for different AI models (like those from Anthropic or OpenAI) and the applications that use them.

Think of it as a universal adapter for AI. Instead of writing custom code to handle the specific API of each AI provider, you can connect to an MCP-compliant server that handles those details for you.

The key benefits of this approach are:

  • Provider Agnostic: You can switch between different AI providers with minimal code changes, often just by changing a parameter in the client.
  • Standardized Tool Use: MCP defines a standard way for AI models to request the use of external tools (like a weather API, a calculator, or a database). Your application can expose these tools to the AI in a consistent way.
  • Simplified Communication: The protocol handles the complexities of streaming responses, managing conversation history, and handling different data formats.

How mcpconn Helps

While MCP provides the standard, mcpconn provides the convenience. It acts as a high-level client library that abstracts away the low-level details of the protocol.

Instead of manually constructing MCP messages, you can use the intuitive methods on the mcpconn:

  • client.connect(): Handles establishing the connection over different transports (like STDIO or HTTP).
  • client.query(): Sends a message to the AI and automatically handles the back-and-forth of tool usage.
  • client.start_conversation(): Manages session IDs and history.

By using mcpconn, you can focus on building your application's logic instead of worrying about the intricacies of AI integration.