SpecterOps/skills

What it solves

This project provides a collection of reusable skills, plugins, and agent definitions designed to streamline security operations, research, and development workflows. It bridges the gap between specialized security tools and AI-driven agent platforms, allowing users to automate complex tasks like reconnaissance, code review, and report drafting.

How it works

The repository contains modular plugins that can be integrated into AI agent environments like Codex or Claude Code. Each plugin includes specific manifests that define its capabilities. For advanced functionality, some plugins support Model Context Protocol (MCP) servers, which allow the agent to interact directly with external security tools like BloodHound, Ghostwriter, or binary analysis software. Users can install these plugins via marketplace commands or use npx to pull in specific skill instructions.

Who it’s for

This project is intended for security professionals, researchers, and developers who use AI agents to assist with offensive security operations, application security assessments, or infrastructure management and want to standardize these workflows.

Highlights

• Extensive library of security-focused plugins covering areas like OSINT, C2 framework development, and reverse engineering.
• Pre-configured agent definitions for specific roles such as "exploit-dev," "report-writer," and "bloodhound-analyst."
• Support for MCP integration to connect agents with live security tooling.
• Modular design allowing for both full plugin installation and standalone skill usage.
• Scaffolding for development workflows, including repository hygiene and security review automation.