MCP Server

A Python - based service application that provides functions for getting metadata from SharePoint and retrieving entities from the Neo4j knowledge graph through semantic similarity search.

Knowledge management and memory Search tools #Metadata retrieval #Knowledge graph #Semantic search #SharePoint .Python

rating : 2 points

downloads : 13

update time : 2025-04-29

What is MCP Server?

MCP Server is an intelligent data retrieval platform that can extract structured metadata from the enterprise SharePoint document library and find relevant information in the knowledge graph through advanced semantic search technology. It is built on the FastMCP framework, providing strong support for knowledge management and information retrieval.

How to use MCP Server?

Users can access the two core functions of MCP Server through simple API calls: getting document metadata and querying the knowledge graph. The system will automatically handle the complex semantic analysis and data retrieval process.

Applicable scenarios

Suitable for scenarios that require the integration of structured and unstructured data, such as enterprise knowledge management, intelligent document retrieval, expert system construction, and decision support systems.

Main features

Metadata retrievalAutomatically extract structured information such as document title, author, and modification date from the SharePoint document library

Semantic searchUse advanced AI models to understand the query intention and find semantically relevant entities and relationships in the knowledge graph

Knowledge graph integrationSeamlessly integrate with the Neo4j graph database and visually display the complex relationship network between entities

Advantages and limitations

Advantages

Intelligent semantic understanding, surpassing traditional keyword - matching search

Unified access to enterprise document and knowledge graph data

Based on the mature FastMCP framework, stable and reliable

Flexible API interface, easy to integrate

Limitations

Requires pre - configuration of SharePoint and Neo4j connections

Semantic search performance depends on the quality of the AI model

Initial deployment requires certain technical knowledge

How to use

Environment preparation

Ensure that Python 3.x is installed and the access permissions for SharePoint and Neo4j are configured

Install dependencies

Use pip to install all necessary Python packages

Configure environment variables

Set the connection parameters for SharePoint and Neo4j in the.env file

Start the service

Run the main program to start the MCP server

Usage examples

Find project documentsEmployees in the marketing department need to find all documents related to the 'Fourth - quarter marketing campaign'

Expert discoveryThe R & D manager is looking for experts with machine learning experience in the company

Frequently Asked Questions

What kind of hardware configuration does MCP Server require?

How to update the data in the knowledge graph?

Does it support other document management systems instead of SharePoint?

Related resources

FastMCP framework documentation

Technical documentation of the core framework on which MCP Server is based

Neo4j graph database guide

Official documentation of the knowledge graph backend database

SharePoint REST API reference

Microsoft's official SharePoint development documentation

🚀 MCP Server

A Python-based server application that offers tools for retrieving metadata from SharePoint and entities from a Neo4j knowledge graph using semantic similarity search.

🚀 Quick Start

This MCP Server is a Python application designed to retrieve metadata from SharePoint and entities from a Neo4j knowledge graph. Here's how to get started:

Prerequisites

Python 3.x
Neo4j database
SharePoint access credentials
OpenAI API key (for semantic search)

Installation

Clone the repository:

git clone <repository-url>
cd mcp-server

Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

Install the dependencies:

pip install -r requirements.txt

Configure environment variables: Create a .env file with the following variables:

SHAREPOINT_URL=
SHAREPOINT_USERNAME=
SHAREPOINT_PASSWORD=
SHAREPOINT_LIBRARY_NAME=

NEO4J_URI=
NEO4J_USER=
NEO4J_PASSWORD=

EMBEDDING_MODEL_ENDPOINT=
EMBEDDING_MODEL_KEY=
EMBEDDING_MODEL_API_VERSION=
EMBEDDING_MODEL_DEPLOYMENT_NAME=
EMBEDDING_MODEL_CHUNK_SIZE=
EMBEDDING_MODEL_DIMENSION=
EMBEDDING_MODEL_CHUNK_OVERLAP=

Usage

Start the server:

python mcp_server.py

The server provides two main tools:
- metadata_retrieve: Retrieve structured metadata from SharePoint.
- entity_retrieve: Query the knowledge graph using semantic similarity.

✨ Features

Metadata Retrieval: Fetch structured metadata from the 'Documents' library in SharePoint.
Entity Retrieval: Query the Neo4j knowledge graph via semantic similarity to find relevant entities and their relationships.
FastMCP Integration: Built on the FastMCP framework for tool management and execution.

📚 Documentation

Project Structure

.
├── mcp_server.py          # Main server application
├── settings.py            # Configuration management
├── services/             # Service implementations
│   ├── graph_store.py    # Neo4j graph operations
│   └── sharepoint.py     # SharePoint integration
├── requirements.txt      # Project dependencies
└── .env                  # Environment variables