Agentic AI Systems

What this domain covers

This edition groups 40 capabilities in Agentic AI Systems across 9 named categories. The inventory contains 25 concepts and 15 tools; 14 skills appeared in at least two of the three original research runs. The remaining entries stay visible with their lower coverage so a reader can distinguish taxonomy scope from research-system agreement.

Current category labels: Agent Applications · Agent Architecture · Agent Control & Oversight · Agent Frameworks · Agent Memory · Agent Protocols · Multi-Agent Systems · Tool Use & Protocols · and 1 more

Frequent learning foundations

1. AI Agent Design supports 14 mapped skills
2. Docker supports 2 mapped skills
3. LangGraph supports 2 mapped skills
4. LLM Function Calling supports 2 mapped skills
5. Multi-Agent Orchestration supports 2 mapped skills