AboutSecurity

Why we included this project

Red teams that already run AI coding agents inside engagements will find this more useful than a static cheatsheet dump. The repo packs 200+ penetration testing methodologies into structured skill files that Claude Code and similar agents can load on their own, so an operator can point an agent at a target and have it pull the playbooks it needs for the current phase, whether that's recon, exploitation, or post-exploitation. It also ships brute-force dictionaries, attack payloads for common web flaws, and a large catalog of product-specific vulnerabilities with PoC details. The material is organized around real engagement phases, from subdomain enumeration and cloud misconfigurations through lateral movement and persistence, which helps both offensive teams and defenders studying how attacks chain together. It is a knowledge base and methodology collection rather than a standalone tool, so plan to pair it with your own agent harness or the companion MCP service.