ai-security

Detect prompt-injection and unauthorized access vulnerabilities in LLM services.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/hanc00l/nemo-agent --skill ai-security-hanc00l
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ai-security
Source: https://github.com/hanc00l/nemo-agent/tree/main/claude-code/.claude/skills/pentest/ai-security
Command: npx skills add https://github.com/hanc00l/nemo-agent --skill ai-security-hanc00l

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps security professionals discover and mitigate AI‑related vulnerabilities such as prompt‑injection, unauthorized API access, data leakage, and model theft in large language model services.

Core Features & Use Cases

  • Prompt injection detection: Scripts to test LLM endpoints for prompt‑injection vulnerabilities.
  • API endpoint discovery: Automated probing of common LLM API paths to locate active services.
  • Vulnerability enumeration: Checklist of common AI security issues with detection methods.
  • Use case: A penetration tester assessing an AI SaaS offering can run this skill to quickly identify insecure configurations and attack surfaces.

Quick Start

Run the ai-security skill to scan the target LLM endpoint for prompt‑injection and unauthorized API access.

Frequently Asked Questions about ai-security

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

FAQPage Schema
How do I detect prompt injection vulnerabilities in LLM services?

To detect prompt injection vulnerabilities in LLM services, you can run automated scripts that probe API endpoints and evaluate prompt handling to identify insecure configurations and attack surfaces.

What is the best way to probe API endpoints for unauthorized access during AI penetration testing?

The best way to probe API endpoints for unauthorized access during AI penetration testing is to use automated discovery scripts that scan common LLM API paths to locate active services and enumerate security flaws like data leakage.

Do I need an HTTP client library to scan LLM endpoints for security flaws?

Yes, you need an HTTP client library along with the target base URL to perform endpoint probing and evaluate prompt handling when scanning LLM infrastructure for security flaws.

Can I use this approach to find model theft and data leakage vulnerabilities in AI infrastructure?

Yes, you can use this approach to find model theft and data leakage vulnerabilities by running vulnerability enumeration checklists that scan AI services for common security issues and insecure configurations.

What security issues can I enumerate when assessing an AI SaaS offering?

When assessing an AI SaaS offering, you can enumerate security issues including prompt injection, unauthorized API access, data leakage, and model theft by probing active API endpoints and evaluating prompt handling.