AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
sandbox — An isolated environment where AI-generated code or agent actions run without being able to touch anything real.
eval — A repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.
Why it matters
A replacement for evalA repeatable test for AI quality — a set of tasks plus scoring — used the way software teams use test suites, because model output is too variable to judge by eyeballing.Full definition → when agents write programs that call your systems: the sandboxAn isolated environment where AI-generated code or agent actions run without being able to touch anything real.Full definition → has no route to your secrets or internal services, and it can pause mid-program for authentication or human approval and resume where it stopped.
