AI Agents for Cybersecurity
Course materials covering chat models, prompt engineering, prompt chaining, retrieval-augmented generation, and agentic workflows applied to cybersecurity operations, with runnable Python examples for each topic.
This repository holds the code and materials for a series of courses on AI agents and agentic systems for cybersecurity operations. It is organized into five parts that build from foundational AI interactions to advanced agentic implementations: basic chat models and embeddings with a visualization exercise built on CVE data, prompt template design for security contexts, prompt-chaining workflows for incident analysis and threat hunting, retrieval-augmented generation with vector databases and web scraping, and agentic RAG and MCP implementations for security automation.
Examples integrate multiple model providers, including OpenAI, Hugging Face, and Ollama, and use frameworks such as LangChain and LangGraph. A Streamlit chatbot example and a Hugging Face sentiment-analysis example are included alongside the security-focused material.
It is intended as a structured, hands-on learning path rather than a standalone tool.