SpecterOps/skills
A marketplace for LLM skills
What it solves
This repository provides a library of reusable agent skills, plugins, and agent definitions designed to extend the capabilities of AI agents used in security operations and offensive security tradecraft.
How it works
It functions as a marketplace of plugins and skills that can be integrated into agent frameworks like Codex and Claude Code. These plugins include manifests and instructions that define how an agent should perform specific tasks. Some plugins are "MCP-aware," meaning they can connect to Model Context Protocol (MCP) servers to interact with external tools like BloodHound, Ghostwriter, and Binary Ninja.
Who it’s for
Security researchers, penetration testers, and offensive security operators who use AI agents to automate reconnaissance, exploit development, and report drafting.
Highlights
- Broad Security Domain Coverage: Includes plugins for OSINT, Active Directory analysis (BloodHound), reverse engineering, and C2 framework development (Mythic, Cobalt Strike).
- Agent Definitions: Provides pre-configured agent personas such as
exploit-dev,bloodhound-analyst, andsocial-engineer. - Cross-Framework Support: Compatible with Codex, Claude Code, and can be installed via
npx skills. - MCP Integration: Enables agents to interact with specialized security tools via the Model Context Protocol.
Related
- Project
- Project
- Project
- Project