oai-mcp

siddhivinayak-sk/oai-mcp

3.1

If you are the rightful owner of oai-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

This project implements a Model Context Protocol (MCP) server using FastAPI, integrating Azure OpenAI and Azure AI Search for enhanced query responses.

MCP Server with Azure OpenAI and Azure AI Search

Overview

This project implements a Model Context Protocol (MCP) server in Python using FastAPI. It provides a /query endpoint and a custom command query that, when detected in a query, enriches the request with context from Azure AI Search and sends it to Azure OpenAI for a response.

Features

  • Standalone MCP server using FastAPI
  • Dockerized for easy deployment
  • Custom query command triggers Azure AI Search and OpenAI integration
  • Configurable via environment variables

Configuration

Set the following environment variables (see docker-compose.yml for example):

Azure OpenAI

  • AZURE_OPENAI_MODEL: Deployed model name
  • AZURE_OPENAI_VERSION: API version
  • AZURE_OPENAI_ENDPOINT: Endpoint URL
  • AZURE_OPENAI_KEY: API key

Azure AI Search

  • AZURE_SEARCH_ENDPOINT: Endpoint URL
  • AZURE_SEARCH_KEY: API key
  • AZURE_SEARCH_INDEX: Index name

Installation

Local (Python 3.10+ required)

pip install -r requirements.txt
uvicorn mcp_server.main:app --reload

Docker

docker-compose up --build

Usage

uv run oai-mcp-server --transport sse --port 8000

Send a POST request to http://localhost:8000/query with JSON body:

{
  "query": "query What is the latest update?"
}

If the query contains query, the server will use Azure AI Search and OpenAI to answer.

Extending

  • Add more commands by extending the FastAPI app and handler modules.

License

MIT