skills.rest
jinyimeng01
jinyimeng01COMMUNITY
Explore all skills by this author →
GitHub

ai-security

Evaluate LLM applications for prompt injection, jailbreaks, and unsafe tool usage vulnerabilities.

54|5|Updated May 9, 2026
Search Tags:#llm#rag#jailbreak#ai security#prompt injection#adversarial testing#agent security
One-click install
npx skills add https://github.com/jinyimeng01/mastermind-bug-bounty --skill ai-security-jinyimeng01
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ai-security
Source: https://github.com/jinyimeng01/mastermind-bug-bounty/tree/main/skills/ai_security
Command: npx skills add https://github.com/jinyimeng01/mastermind-bug-bounty --skill ai-security-jinyimeng01

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps security professionals identify and evaluate vulnerabilities in AI and LLM applications, including prompt injection, jailbreaks, data leakage, and unsafe tool usage.

Core Features & Use Cases

  • AI Attack Testing: Provides structured methodologies and test vectors for prompt injection, role escape, system prompt extraction, and adversarial inputs.
  • LLM Security Assessment: Covers RAG poisoning, MCP and agent attacks, tool/function call abuse, and model extraction scenarios.
  • Use Case: A security researcher can use this Skill to build a comprehensive AI application assessment plan and validate whether an LLM-based product resists common attack techniques.

Quick Start

Use the ai-security skill to evaluate an LLM application for prompt injection, jailbreak, and tool abuse vulnerabilities.

Frequently Asked Questions about ai-security

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

FAQPage Schema
How do I test LLM applications for prompt injection and jailbreak vulnerabilities?▼

To test LLM applications for prompt injection and jailbreaks, apply structured security test vectors and methodologies that evaluate model interactions, role escape attempts, and system prompt extraction.

What is RAG poisoning and how does it affect AI agent security?▼

RAG poisoning is an adversarial attack corrupting retrieval data to manipulate AI outputs. It compromises AI agent security by inducing unsafe tool usage, function call abuse, and unauthorized data exposure.

How do I build a comprehensive AI application assessment plan for model extraction scenarios?▼

Build a comprehensive AI application assessment plan by defining structured test vectors for model extraction, analyzing tool usage risks, and validating resistance against common LLM attack techniques.

Does this AI security testing approach cover MCP and agent attacks?▼

Yes, this AI security testing approach covers MCP and agent attacks. It evaluates tool and function call abuses, RAG poisoning, and model extraction scenarios within LLM application assessments.

What are the limitations of automated adversarial testing for unsafe model behaviors?▼

Automated adversarial testing for unsafe model behaviors is limited by the need for structured test vectors. It requires continuous methodology updates to validate evolving model interactions and data exposure risks.

Related Skills

Explore More Software Engineering Skills→
alexmaruccialexmarucci

thinking-red-team

Analyze security plans and architectures for attack vectors and weaknesses.

Community
Advanced
tiankong0101-bytetiankong0101-byte

moltguard

Detect security vulnerabilities and configuration risks in agent deployments and OpenClaw environments.

Community
Advanced
kslhunterkslhunter
2

sd-explore

Trace code execution paths from entry points to data storage.

Community
Advanced
ContextJet-aiContextJet-ai
29

red-team-llm-app

Red-team LLM applications against security vulnerabilities like prompt injection and jailbreaks.

Official
Advanced