red-teaming

Plan and execute red team operations against cybersecurity and AI/LLM systems.

100|15|Updated Nov 13, 2025

Search Tags:#penetration testing #cybersecurity #ai security #prompt injection #llm security #red teaming #adversary emulation

Ready to run

npx skills add https://github.com/rfxlamia/skillkit --skill red-teaming-rfxlamia

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: red-teaming
Source: https://github.com/rfxlamia/skillkit/tree/main/skills/red-teaming
Command: npx skills add https://github.com/rfxlamia/skillkit --skill red-teaming-rfxlamia

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides a comprehensive methodology for simulating real-world attacks against both traditional IT infrastructure and AI/LLM systems, enabling robust security validation and adversary emulation.

Core Features & Use Cases

Dual-Domain Red Teaming: Covers cybersecurity (network, system compromise) and AI/LLM security (prompt injection, safety testing).
Structured Methodology: Follows a phased approach (Planning, Reconnaissance, Execution, Reporting) aligned with industry frameworks like MITRE ATT&CK and OWASP LLM Top 10.
Use Case: A company wants to test its new AI chatbot's resistance to prompt injection and its overall security posture against common cyber threats. This Skill can guide the entire process, from planning attack scenarios to generating a final report with actionable remediation steps.

Quick Start

Use the red-teaming skill to plan a cybersecurity red team exercise targeting our web application.

Frequently Asked Questions about red-teaming

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I plan a red team exercise against an LLM chatbot?▼

Red teaming AI systems involves simulating attacks like prompt injection to validate LLM safety. This methodology provides a phased approach to test resistance against adversarial threats, ensuring your chatbot maintains a robust security posture.

How does adversary emulation work for cybersecurity validation?▼

Adversary emulation simulates real-world cyber attacks to validate your infrastructure's security posture. By leveraging the MITRE ATT&CK framework, it guides the process from attack execution to reporting, identifying vulnerabilities through structured threat simulation.

Can I test both network infrastructure and AI prompt injection using the same methodology?▼

Yes, you can test both network infrastructure and LLM prompt injection using this dual-domain methodology. It integrates traditional cybersecurity adversary emulation with AI security validation, covering diverse threat landscapes within a single structured approach.

What frameworks are used for LLM security testing and vulnerability assessment?▼

LLM security testing and vulnerability assessment leverage the OWASP LLM Top 10 and MITRE ATT&CK frameworks. These guide the structured methodology for attack execution and reporting, validating defenses against diverse cybersecurity and AI threat landscapes.

How do I generate a security remediation report after a red teaming exercise?▼

To generate a remediation report, you complete the final reporting phase of the red teaming methodology. It translates attack execution and vulnerability assessment results into actionable steps, validating defenses against simulated cybersecurity and AI threats.