hello-agents

Why we included this project

This is one of the most complete open, free curricula we have seen for learning to build LLM agents, and it earns a place here for its depth and hands-on structure rather than as a tool you deploy. It starts with agent fundamentals and LLM basics, moves through classic patterns like ReAct and Reflection, then into low-code platforms and mainstream frameworks, and ends with building your own agent framework on the OpenAI native API. Later chapters cover memory and retrieval, context engineering, communication protocols such as MCP and A2A, and even Agentic RL training from SFT to GRPO, so it goes well beyond a beginner primer. The accompanying code folder and end-to-end projects like a travel assistant and a simulated cyber town give you something to build along with, which matters if you want to move from using LLMs to constructing agent systems. It is a learning resource rather than production software, but for anyone planning to build agents seriously, the structured progression and community extras are genuinely useful.