What is an agent?

useful canonical explanation for separating LLM, tool call, harness, round, and agent; good onboarding piece for people confused by agent jargon.

Original source

What it is: Karthik S sharing Hadley Wickham’s “What is an agent?” explainer.

Gist: very clear mental model: an agent is an LLM inside a harness that can call tools repeatedly in a loop; the harness mediates tool calls/results and turns a stateless request/response model into iterative action.

Newsletter angle: “agent = looped tool use inside a harness” is a concise definitional anchor for broader agent discussions.

Note: extracted tweet via FXTwitter and fetched linked Substack article successfully.