Single Agent
A single LLM receives a user request, reasons over available tools, executes them, and returns a final response. It loops until the task is complete — no coordination with other agents required.
A single agent is the foundational building block of agentic AI — an LLM given a goal, a set of tools, and a loop that lets it act until the task is done. Unlike a simple chatbot, an agent decides which tools to call, processes the results, and keeps reasoning until it reaches its objective. Building one from scratch reveals exactly how the perceive-think-act cycle works in practice.
Ask anything about Single Agent