license: apache-2.0
language: [en, tr]
task_categories: [text-classification]
tags: [ai-security, agent-security, prompt-injection, vulnerable, training, red-teaming, mcp]
pretty_name: "Agent Skill Vulnerability Scenarios"
size_categories: [n<1K]
configs: [{config_name: scenarios, data_files: scen.jsonl}]
Agent Skill Vulnerability Scenarios (defanged, teaching)
from datasets import load_dataset
ds = load_dataset("fevziegeyurtsevenler/agent-skill-vulnerabilities")
Deliberately-vulnerable, defanged agent-extension artifacts — for training detectors & hands-on learning.
Schema
| column | meaning |
|---|---|
| id | scenario |
| artifact_type | SKILL.md / mcp.json |
| content, walkthrough | artifact + defense |
Related AltaySec resources
- 🕵️ uncloak scanner: https://github.com/fevziegeyurtsevenler/uncloak · live demo
- 📚 Other datasets: detection-rules · glossary · skills-in-the-wild
- 🌐 AltaySec · Açık Kaynak Lab
Citation
@misc{yurtsevenler2026agentskillvulnerabilities,
title = {Agent Skill Vulnerability Scenarios},
author = {Yurtsevenler, Fevzi Ege},
year = {2026},
publisher = {AltaySec / Hugging Face},
howpublished = {\url{https://huggingface.co/datasets/fevziegeyurtsevenler/agent-skill-vulnerabilities}}
}
Multilingual-first LLM/agent security research by AltaySec.