Model Context Protocol (MCP)

What is Model Context Protocol (MCP)?

Model Context Protocol (MCP) is a standardized framework that allows developers to connect large language models (LLMs) with external data sources and tools (functions). This repository provides a demo MCP server implementation using streamable HTTP transport, an auto‑tool registry, and a Dockerfile for scalable, production‑ready deployment. It is aimed at developers building AI applications that need to give LLMs access to live data or custom actions.

How to use Model Context Protocol (MCP)?

Install uv (the Astral package manager), then run uv sync to install dependencies. The server can be started locally or inside Docker and deployed to any cloud provider. Tools are registered automatically using the @mcp_tool decorator. A companion MCP client that works with the OpenAI SDK is included in the client/ directory.

Key features of Model Context Protocol (MCP)

• Stateless MCP server using streamable HTTP transport
• Auto‑tool registration via @mcp_tool decorator
• Dockerfile for containerized deployment
• Can be run locally or in any cloud provider
• Demo client integrated with OpenAI SDK

Use cases of Model Context Protocol (MCP)

• Give an LLM access to a live database or API endpoint
• Build a custom assistant that fetches real‑time information
• Combine multiple tools into a single, standardized LLM backend

FAQ from Model Context Protocol (MCP)

What is MCP?

MCP is a protocol that standardizes how LLMs connect with external data sources and tools (functions).

How do I set up the server?

Install uv, run uv sync to install dependencies, then start the server locally or via Docker.

Can I deploy the server to the cloud?

Yes. The server is stateless and can be deployed to any cloud provider using the provided Dockerfile.

How are tools registered?

Tools are automatically registered with the @mcp_tool decorator placed on the tool function.

What transport does the server use?

The server uses streamable HTTP transport, suitable for scalable production deployments.