Content
# Rootkit Matrix MCP Server
A Model Context Protocol (MCP) server built with FastMCP to feed your AI model technical data about rootkit techniques, rootkit and bootkit families from the [Rootkit Techniques Matrix](https://aibaranov.github.io/rootkit-matrix/). Consists of two files: `server.py` and `matrix_data.json`. The latter stores data from the matrix in a format understandable for the methods in `server.py`.
It's compatible with Open WebUI and Claude Windows app. Since I don't have access to Claude, I tested it on a locally deployed model Ollama + Qwen3:8b. But compatibility with Claude was added as well. I don't use Linux, so everything was tested on Windows.
## Installation
1. **Clone the repo**
```cmd
git clone https://github.com/ArtemBaranov/rkmatrix-mcp-server.git
```
2. **Install FastMCP**:
```cmd
pip install fastmcp
```
## Configuration for Ollama & Open WebUI
The server itself doesn't need configuration. You simply run it as `python server.py`. If you are running a local model setup (Ollama + Open WebUI), follow these steps to connect Open WebUI to your MCP server (Ollama app itself can't be directly connected to the MCP server).
1. Install Ollama app and choose (download) `qwen3:8b` model.
2. Now we need to install the Open WebUI framework and connect it to the Ollama service. But since Open WebUI is only compatible with Python 3.11, we need to install [Docker Desktop for Windows](https://www.docker.com/products/docker-desktop/) to create a virtual environment for it. Once installed, run the following command to create the environment.
```
docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main
```
3. Run the container or make sure it's running.
4. Run the server:
```cmd
python server.py --http
```
5. Now, we can setup the server. Go to `http://localhost:3000/`. Navigate to: Your profile icon -> Admin panel -> Settings -> Integration -> External Tool Servers -> Add connection.
* Type: MCP (Streamable HTTP)
* Name: rkmatrix
* ID: rkmatrix
* URL: `http://host.docker.internal:8000/mcp`
* Auth: None
To ensure the model utilizes the MCP tools correctly, change "Function Calling" setting of the model to Legacy. Admin Panel -> Settings -> Models -> Change -> Advanced Params.
Now you should see the toolbox icon near your prompt text box, confirming our tool is ready.
## Configuration for Claude Windows app
To use this server with the Claude Desktop app, add the following configuration to your `claude_desktop_config.json` file:
* **macOS:** `~/Library/Application Support/Claude/claude_desktop_config.json`
* **Windows:** `%APPDATA%\Claude\claude_desktop_config.json`
```json
{
"mcpServers": {
"rkmatrix-mcp-server": {
"command": "python",
"args": [
"/absolute/path/to/your/rkmatrix-mcp-server/server.py"
]
}
}
}
```
*Note: Replace `/absolute/path/to/your/` with the actual full directory path on your machine.*
When using Claude Desktop, the app launches the server automatically in the background. You do not need to manually run `python server.py` in your command prompt.
