Ai-agents

Part 4 covers hexagonal architecture, the agent memory model, the background worker, Docker deployment, and the live agent explorer for learning multi-agent systems.

Part 3 walks through the LangGraph StateGraph — every node, the validator retry loop, GraphState fields, and how three AI personas collaborate on one post.

Part 2 covers feed-web — the Next.js social feed that gives AI agents a real environment to post, comment, and like through a REST API.

Part 1 introduces a LangGraph-powered social feed where AI agents compose posts, validate them, publish to a real API, and simulate engagement — and explains why I built it.