Agent skill · security · aradotso
ai-security-knowledge-base
Comprehensive AI security knowledge base covering ML algorithms, threat modeling, offensive AI tactics, and defensive strategies including OWASP LLM Top 10 and adversarial ML
Why this skill is useful
Includes detailed documentation on AI security vulnerabilities and offensive tactics, along with executable examples for automated vulnerability scanning and adversarial ML evasion.
What it needs
Requires git installed locally. About 6k tokens when loaded. Last updated 2026-08-04. 11 stars on the source repository.
What this skill does
AI Security Knowledge Base Skill Skill by ara.so — Security Skills collection. Overview The AI Security Knowledge Base (AISecurityTop) is a comprehensive Chinese-language security documentation project that covers the complete spectrum of AI security from foundational machine learning algorithms to advanced threats like prompt injection, adversarial attacks, model evasion, and deepfake exploitation. This project serves as both a reference guide and practical resource for security researchers, penetration testers, and AI developers. Core Coverage Areas AI Fundamentals: Deep learning architectures, supervised/unsupervised learning, reinforcement learning Threat Modeling: OWASP ML/LLM Top 10, MCP security, skill security frameworks Red Team Operations: Offensive AI tactics, adversarial ML, deepfake generation, AI-powered penetration testing Blue Team Defense: ML-based threat detection, automated auditing, defensive algorithms Installation This is a documentation/knowledge repository, not a code library. Clone it for reference: Key Documentation Structure 1. Foundational AI Concepts AI.md - Artificial Intelligence overview, history, and core concepts 深度学习.md - Deep learning architectures (CNN, RNN, Transformer) 监督学习算法.md - Supervised learning: Linear/Logistic Regression, SVM, Decision Trees, Random Forests, Naive Bayes 无监督学习算法.md - Unsupervised learning: K-Means, GMM clustering, association rules 强化学习.md - Reinforcement learning: Agent-based decision making and game theory 2. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills aradotso/ai-security-knowledge-base