Answers are grounded in this paper and cite their sources.
This paper looks at AI agent systems as four planes: Intelligence, Orchestration, Tools, and Information. It then defines the workflows that run across those planes, including the kernel workflow, memory, and intelligence call control. It is written for developers and architects who build production agents. It is a way of looking at agent systems, not a specification.
No takes yet. Share an insight, caveat, or question.
Sooraj Kadnor (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: