operator console
A term earns its place by changing the work.
The console is the practical layer of LLM Term. It treats vocabulary as an operating surface: a phrase should help someone write a clearer prompt, review a model answer, explain a failure, or publish a page that machines can parse accurately. Words that only sound impressive are left outside the desk.
scope
Does the term name a bounded task, artifact, or behavior?
signal
Does it help a reader predict what evidence matters?
failure
Can the wording expose a wrong answer or weak instruction?
reuse
Can another team carry the phrase into a prompt or spec without distortion?
How console notes are read
A console note starts with the working situation, then names the term, the common misuse, and the cleaner replacement. For example, “hallucination” is not a blanket synonym for every bad output; sometimes the better term is retrieval miss, stale source, unsupported inference, formatting drift, or instruction conflict. That distinction matters because each failure asks for a different fix. The console format keeps that distinction visible before teams turn a dramatic label into an unhelpful ticket.