zebbern-kali

The Zebbern Kali MCP Server provides an API-enabled gateway to a comprehensive Kali Linux pentesting toolkit. It exposes 139 MCP tool functions and a wide range of external tools (Nmap, SQLMap, Hydra, Metasploit, Nuclei, and more) through a Flask-based API, enabling an AI assistant client (like VS Code with Copilot) to orchestrate security assessments remotely. To begin, ensure your MCP client is configured to connect to the server at the Kali VM's address and port (default 5000). You can health-check the service via the /health endpoint and invoke specific tools via the /api/tools/<tool-name> endpoints with JSON payloads describing targets and parameters. The community documentation outlines tool references, workflows, and API details to help you compose commands for recon, vulnerability assessment, exploitation, password cracking, and AD enumeration. The server also supports session management for Metasploit and reverse shells, SSH connections, and evidence collection (screenshots, notes, findings) during engagements.

Prerequisites:

• A Kali Linux VM (or Debian-based system) with pentest tools installed
• Python 3.8+ and pip
• Git

Option A: Shell script installation (recommended for Kali)

1. Clone the repository and install:

# Clone the repository
git clone https://github.com/zebbern/zebbern-mcp.git
cd zebbern-mcp

# Run the installer (requires root)
sudo ./install.sh

2. The installer sets up:

• Flask API server at /opt/zebbern-kali
• A systemd service named kali-mcp (auto-start on boot)
• All pentesting tools (nmap, sqlmap, nuclei, etc.)

Option B: Python installer

1. Install using Python script (server components):

sudo python3 install.py --server

2. Follow prompts to configure connectivity to the Kali host.

Option C: Remote installation (one-command)

1. From your host machine, install on Kali via SSH:

python install.py --remote --host <KALI_IP> --user kali --password <PASSWORD>

Notes:

• The server component exposes a Flask API and a systemd service named kali-mcp.
• After installation, the server runs the MCP services and integrates tools on the Kali host.

Tips and considerations:

• The server exposes powerful pentesting tools; restrict access to trusted networks and enable authentication where possible.
• DEFAULT_KALI_SERVER should point to the Kali API URL (e.g., http://<kali-ip>:5000).
• Use the provided VS Code MCP config to connect the MCP client; edit the path to mcp_server.py and the Kali server URL as needed.
• If tools are missing, re-run installation with the appropriate flags (e.g., --no-service to skip service setup, --tools to install tools only).
• For troubleshooting, check the kali-mcp systemd service status and logs via journalctl, verify port 5000 openness, and ensure the Python environment is intact.
• Security: This server provides unrestricted access to powerful tools; run in isolated networks and enforce access controls.