AutoFortify

AutoFortify is a suite of Model Context Protocol (MCP) tools that empowers IT administrators to secure Windows devices and services using natural language.

Features

• Natural Language Security Management: Interact with Windows security tools using LLMs and context
• Real-time Chat Interface: User-friendly web interface powered by Chainlit
• MCP Tools: Extensible architecture using Model Context Protocol
  • Windows Firewall
  • Active Directory (requires Windows Server AD DC)
  • Bloodhound (with additional capabilities)

Quick Start

Prerequisites

• Python 3.12+
• Windows 10/11/Server (for Windows Security features)
• Azure OpenAI API access
• BloodHound Community Edition (optional, for AD analysis)

Installation

1. Clone the repository

   git clone https://github.com/AutoFortify/autofortify.git
   cd autofortify

2. Set up the Agent

   cd agent
   pip install -r requirements.txt

3. Configure environment variables

   Create .env with your Azure OpenAI credentials:

   AOAI_ENDPOINT_URI=https://your-openai-endpoint.cognitiveservices.azure.com/
   AOAI_API_KEY=your-api-key
   AOAI_API_VERSION=2025-03-01-preview
   MCP_SERVER_URL=http://127.0.0.1:8081/mcp
   

4. Start the MCP Server

   cd ../mcp
   pip install -r requirements.txt
   python main.py

5. Start the Agent Interface

   cd ../agent
   chainlit run app.py -w -h

Architecture

The MCP servers are built using the FastMCP library. We also implement an MCP client and agent using Semantic Kernel, while Chainlit provides the frontend for the agent.

Project Structure

autofortify/
├── agent/                # Optional MCP client
│   ├── app.py            # Chainlit + Semantic Kernel MCP client
└── mcp/                  # MCP servers/tools
    ├── main.py           # Windows Firewall + Active Directory MCP server
    └── bloodhound/       # BloodHound MCP server

Disclaimer

Caution

This code is NOT intended for production use, and is only intended to be run as a demo in a local, virtualized environment.