Category · 6 terms
AI Agents & Autonomy
defined clearly.
Tool-using, planning, and multi-step systems that act on their own.
AI Agent
An LLM-powered system that perceives its environment, plans actions, invokes tools, and pursues goals autonomously across multiple steps.
Function Calling
OpenAI's term for tool use — when a model emits structured JSON specifying which function to call and with what arguments.
MCP (Model Context Protocol)
Anthropic's open standard for connecting LLMs to external tools and data sources via a standardised server protocol.
Multi-Agent System (MAS)
A system of multiple interacting AI agents that collaborate, compete, or coordinate — the architectural pattern behind complex LLM-based workflows.
ReAct (Reason + Act)
An agent design pattern that interleaves explicit reasoning ("thought") with actions and observations, making the agent's logic auditable.
Tool Use (Function Calling)
The capability for an LLM to invoke external functions or APIs to retrieve information or perform actions — the foundation of agent architectures.