Call My "Agent"
The word “agent” have been used with greater frequency over the last few years. I seemed to have started using this word without a deep understanding of the overarching concept. This article aims to provide an exploration of this term.
The Beginning
We will take the above model for the LLM interface as a given (to delve into this black box, see Internals of a Large Language Model (LLM)). This interface can be enhanced by augmenting capabilities (“Tools”) to inspect and mutate content. One mechanism for the LLM to access these capabilities is via the Model Context Protocol (MCP).
Stringing several of these calls together, with a verification step that decides whether to iterate or stop, is the rough shape of an “agentic” loop. Press the check to keep iterating, or the cross to fail out and restart.