Community-maintained skills for offensive and defensive security capabilities. Built for security researchers, engineers, and bug hunters.
Skills extend AI agents with specialized security capabilities. When enabled, skills provide domain-specific instructions, scripts, and references that help agents perform complex security tasks.
This repository follows the Agent Skills specification.
This repo is also a Claude Code plugin marketplace. Each skill is installable as a plugin.
# Register the marketplace
/plugin marketplace add HacktronAI/skills
# Browse and install from the menu
/plugin menu
# Or install a specific skill directly
/plugin install finding-triage@hacktron
Install from a local clone instead:
git clone https://github.com/HacktronAI/skills.git# From the parent directory of the clone
/plugin marketplace add ./skills
/plugin install finding-triage@hacktron
# Pull the skills registry
hacktron skills pull
# List available skills
hacktron skills list
# Enable a skill
hacktron skills enable patch-diff-analyzer
# Disable a skill
hacktron skills disable patch-diff-analyzerSkills are automatically discovered and can be enabled/disabled from the extension settings.
| Skill | Description | Sources |
|---|---|---|
| finding-triage | Interactively triage Hacktron findings against source (and optionally a live deployment), then fix + commit confirmed issues or set their state in Hacktron | - |
| patch-diff-analyzer | Reverse-engineer compiled binaries (JARs, DLLs) to analyze security patches | - |
Skills from other repositories that follow the Agent Skills specification can be used with Hacktron:
- SecOpsAgentKit - 25+ security operations skills including SAST, DAST, container scanning, and secret detection
- Raptor Skills - Additional security research skills
To use skills from other repositories, manually copy them into ~/.hacktron/skills/:
# Copy individual skills from other repos
cp -r /path/to/other-repo/skill-name ~/.hacktron/skills/Security Warning: Skills can execute arbitrary commands on your machine. Always review the
SKILL.mdand any scripts before adding skills from third-party sources. Only the official HacktronAI/skills repository is reviewed and validated for security.
Note: The
hacktron skills pullcommand only pulls from the official HacktronAI/skills repository. Third-party skills must be manually copied to ensure users consciously review what they're installing.
Skills follow the Agent Skills specification. Each skill is a directory containing:
skill-name/
├── SKILL.md # Required - YAML frontmatter + instructions
├── scripts/ # Optional - executable scripts
├── references/ # Optional - additional documentation
└── assets/ # Optional - templates, data files
---
name: skill-name
description: What the skill does and when to use it.
license: MIT
compatibility: Required tools or environment
metadata:
author: your-name
version: "1.0.0"
---
# Skill Name
Instructions for the agent...- Fork this repository
- Create your skill following the spec
- Test with
hacktron skills enable your-skill - Open a PR
All skills are reviewed for security before being merged.
MIT