Agents are the productization of LLMs - the layer that turns 'interesting chatbot' into 'software that actually does things for you.' This 90-page knowledge base covers the full stack: foundations (what makes an agent an agent), frameworks, the canonical patterns, tool use, memory, multi-agent orchestration, evaluation, and production deployment. It pairs naturally with the Framework section on this site (which is Claude-specific) - this one is more broadly about agents regardless of which model you use. If the future of software is 'it kind of runs itself,' this is where you learn to build it.
What agents are, when to use them, the architecture map.
ReAct, planning, reflection, self-correction.
Tool design, descriptions, error handling, budgets.
Short-term, long-term, episodic, procedural.
Orchestration patterns, handoffs, debate.
Task completion, trajectory eval, regression testing.
Observability, cost, latency, safety, human-in-loop.
Research, coding, support, data, browser, email agents.
Claude Agent SDK, LangGraph, CrewAI, AutoGen.
Start at Foundations. If you already know what agents are, jump to Loops then Tools. Evaluation and Production are where naive agents fall apart at scale.
Andrej Karpathy - Intro to Large Language Models (1 hour)