Provides a bridge to run Kali terminal commands from MCP clients and AI endpoints.
Configuration
View docs{
"mcpServers": {
"wh0am123-mcp-kali-server": {
"url": "http://LINUX_IP:5000/"
}
}
}The Kali MCP Server provides a lightweight bridge that lets MCP Clients run terminal commands on a Kali Linux machine through a controlled API, enabling AI-assisted testing, automated recon, and real-time interaction with web services from your MCP apps.
You use an MCP client to connect to your Kali MCP Server and start issuing terminal commands through the API. The server exposes a programmatic interface that accepts requests from AI endpoints or MCP clients, executes the requested commands on Kali, and returns structured results. You can perform common security testing tasks like network scanning, web requests, and utility-driven workflows, all orchestrated by your MCP client.
Prerequisites: ensure Python is installed on both the Kali machine and any MCP client machines. Install Python packages listed in the project requirements.
On your Kali machine follow these steps to install and run the server.
git clone https://github.com/Wh0am123/MCP-Kali-Server.git
cd MCP-Kali-Server
pip install -r requirements.txt
python3 kali_server.pyClone the MCP Kali server project on the client as a starting point, install dependencies, and connect to the Kali MCP server using the provided client command.
git clone https://github.com/Wh0am123/MCP-Kali-Server.git
cd MCP-Kali-Server
pip install -r requirements.txtIf both the client and Kali server run on the same machine, start the client with the server address for the Kali host.
If the client and Kali server are on different machines, establish an SSH tunnel to forward the Kali MCP server port to the client machine, then run the client pointed at the tunneled address.
# Example SSH tunnel and client start
ssh -L 5000:localhost:5000 user@KALI_IP
./mcp_server.py --server http://127.0.0.1:5000The following configuration shows how an MCP client can connect to the Kali MCP server using both HTTP and stdio methods. Use the method your client prefers. The HTTP method uses a remote URL, while the stdio method runs a local process that connects to the server.
{
"type": "http",
"name": "kali_mcp_http",
"url": "http://LINUX_IP:5000/",
"args": []
}{
"type": "stdio",
"name": "kali_mcp",
"command": "python3",
"args": ["/absolute/path/to/mcp_server.py", "--server", "http://LINUX_IP:5000/"],
"env": []
}Connect your Kali MCP Server to AI endpoints (OpenAI, Claude, DeepSeek, etc.) to drive terminal command decisions.
Expose a controlled API to execute terminal commands on the Kali Linux machine.
Allow AI to interact with websites and APIs, capture flags via curl or other tools, and automate tasks.
Designed for red team activities, bug bounties, and CTF challenges with automated workflows.