Skip to content
Get packs

Guide · Open source

How to use HexStrike AI: install, connect an AI client and run it

Updated · written with AI from the project's own documentation and checked against our catalogue data

HexStrike AI is an open-source MCP server, written in Python, that lets AI assistants such as Claude, GPT and Copilot run a large set of existing cybersecurity tools. This guide walks through what it needs, how to install and start it, how to connect an AI client, and what to check before using it. It is written for readers who already have Linux and security tooling experience, since the README assumes that.

HexStrike AI0x4m4/hexstrike-ai · ★ 13.3k · MIT

Quick start

  1. 1Clone the repository. Run git clone on the 0x4m4/hexstrike-ai repository and change into the folder.
  2. 2Create a virtual environment. Run python3 -m venv hexstrike-env and activate it with source hexstrike-env/bin/activate.
  3. 3Install Python dependencies. Run pip3 install -r requirements.txt inside the activated environment.
  4. 4Install the external tools you need. Install the security tools and, if needed, Chromium or Chrome, from your Linux package sources.
  5. 5Start the MCP server. Run python3 hexstrike_server.py and check http://localhost:8888/health.
  6. 6Connect your AI client. Add the hexstrike_mcp.py entry with the server address to your client's MCP configuration and restart the client.

What HexStrike AI does and who it is for

HexStrike AI is an MCP (Model Context Protocol) server. An AI agent connects to it and can then run cybersecurity tools such as scanners, subdomain finders and password crackers, and chain their results together for penetration testing, vulnerability discovery and security research. The project is owned by OTT Cybersecurity LLC and released under the MIT license.

  • Security professionals who want to script or speed up recon with an AI assistant
  • Students practicing on CTF challenges or lab machines they own
  • Researchers experimenting with how LLM agents use security tooling
  • Developers who already run Kali-style tool setups and want an MCP bridge

Check the prerequisites before you start

The README lists Python 3.8 or newer, pip3, and a Linux environment such as Kali, since many of the listed tools are Kali-oriented. If you want the Browser Agent, you also need Chromium or Google Chrome.

  • Python 3.8 or newer
  • pip3
  • A Linux environment such as Kali
  • Chromium or Google Chrome for the Browser Agent
  • Only test systems you own or have written permission to test

Install HexStrike AI from source

HexStrike AI is installed by cloning the repository, creating a virtual environment and installing the Python dependencies. Run these commands from a terminal in the directory where you want the project. The README's Linux and Mac activation line is shown here; on Windows it lists a different activation path.

  • Clone the repository, then change into the hexstrike-ai folder
  • Create and activate a virtual environment named hexstrike-env
  • Install the packages listed in requirements.txt with pip3
git clone https://github.com/0x4m4/hexstrike-ai.git cd hexstrike-ai python3 -m venv hexstrike-env source hexstrike-env/bin/activate pip3 install -r requirements.txt

Install the external security tools

The Python packages do not include the security tools themselves. The README lists core network, web, password and binary analysis tools, plus cloud security tools, that you install separately. It does not give one command that installs all of them, so expect to install and troubleshoot them one group at a time. Chromium or Chrome is also needed for the Browser Agent, and the README gives apt commands for it.

  • Core tools include nmap, masscan, amass, subfinder, nuclei, gobuster, ffuf, sqlmap, hydra, john and hashcat
  • Cloud tools include prowler, scout-suite and trivy
  • Browser Agent needs chromium or google-chrome-stable
sudo apt install chromium-browser chromium-chromedriver

Start the server and verify it is running

Start the MCP server from the project folder with python3 hexstrike_server.py. The README lists two optional flags: --debug for debug mode and --port to choose a port. The examples use port 8888. A health check and a sample analysis request are shown in the README.

  • Default start: python3 hexstrike_server.py
  • Debug mode: add --debug
  • Custom port: add --port 8888 (the README's example port)
python3 hexstrike_server.py curl http://localhost:8888/health

Connect an AI client to the server

The README shows configuration examples for Claude Desktop and VS Code Copilot. Both point the client at the hexstrike_mcp.py script and pass the server address. Replace /path/to/hexstrike-ai with the real location of your clone. The README also names 5ire (noting v0.14.0 is not supported for now), Roo Code and Cursor, and refers to a video for those.

  • Claude Desktop: edit ~/.config/Claude/claude_desktop_config.json
  • VS Code Copilot: edit .vscode/settings.json
  • Use the server address http://localhost:8888 unless you changed the port
{"mcpServers":{"hexstrike-ai":{"command":"python3","args":["/path/to/hexstrike-ai/hexstrike_mcp.py","--server","http://localhost:8888"]}}}

What people use it for

  • Scripting a recon pass with an AI assistant that chains subdomain discovery into port scanning and web directory enumeration on a lab machine you own
  • Working through CTF challenges with an AI agent that calls local tools like binwalk, steghide, exiftool and strings
  • Studying how LLM agents select and sequence security tools, using the README's description of tool selection and attack chain discovery
  • Running the Browser Agent for web tasks on a lab target, once Chromium or Chrome is installed

Frequently asked questions

Is HexStrike AI free to use?

The project is released under the MIT license, which the README lists, so the code is open source. The README does not mention any paid tier. Check the LICENSE file in the repository for the full terms.

What are the requirements for HexStrike AI?

The README lists Python 3.8 or newer, pip3, and a Linux environment such as Kali. Chromium or Google Chrome is needed only for the Browser Agent. Many of the listed security tools must be installed separately.

Which AI clients does HexStrike AI work with?

The README names Claude Desktop, VS Code Copilot, Roo Code, Cursor and 5ire, and says any MCP-compatible agent can connect. It notes that 5ire version v0.14.0 is not supported for now.

Why won't HexStrike AI start or find a tool?

The README does not include a troubleshooting section. Start by confirming Python 3.8 or newer is active inside the virtual environment, that requirements.txt installed without errors, and that the external tool is on your PATH. Run the server with --debug to see more output.

What are the alternatives to HexStrike AI?

The comparable projects the directory lists are PentestGPT, Metasploit Framework, Nuclei and Burp Suite Community. HexStrike AI differs by letting an AI agent drive many existing tools through MCP, whereas those projects are used directly by the tester.

Is HexStrike AI legal to use?

The tools are offensive, and the README does not address legal scope. Only test systems you own or have written permission to test, such as your own lab machines or CTF challenges.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

HexStrike AI suits security professionals and students who run a Linux and Kali-style setup, and who want an AI agent to drive existing tools on systems they are authorised to test. It is a poor fit for beginners, for anyone without written permission to test a target, and for teams that need a supported, polished product, since the README is largely promotional and no release is tagged.

Install options & downloads