NUILab Agentic AI
NUILab Agentic AI
Concepts
What agentic AI is, and the core ideas behind it — agents, tools, memory, planning, and autonomy.
What agentic AI is
A plain-language take on what makes an AI system 'agentic', and how it differs from a chatbot.
ConceptsA multi-machine agent stack
Run coding and research agents across several machines at once — in parallel, able to coordinate, mixing local and frontier models — and stay in control of what they do.
ConceptsSubscriptions, APIs, and local models
Three ways to get a model behind your agent — a flat-rate subscription, a pay-per-token API, or a model on your own hardware — and which client each one requires.
ConceptsSkills, MCP, and extending an agent
Beyond standing instructions: packaged skills the agent loads when relevant, MCP servers that connect it to your tools and data, and the commands, subagents, and hooks around them — with the cross-vendor equivalents.