hexstrike-ai

Why we included this project

Security engineers who want their AI assistants to run real offensive tooling, not just describe it, should look at this project. It bundles more than 150 command-line security utilities, from network reconnaissance to web application testing, OSINT, and binary analysis, behind the Model Context Protocol. Claude, GPT, Copilot, and similar models can invoke those tools as structured steps inside a pentest or bug bounty workflow. The framework also includes prebuilt agents that chain tools together and track state, so an operator does not have to juggle terminals and manually pass output between utilities. Tooling and agent logic ship in one installable package.