Agent Loop

You might say

After a check fails, should the AI continue editing, stop to ask me, or end the task?

A runtime mechanism that repeatedly calls AI and tools, then uses check results to continue, pause, or stopWhile fixing horizontal overflow in mobile navigation, the first failed check can start another round. A second identical failure or a change to shared styles pauses for confirmation. The task ends only after the build and 390px page check both pass.
Agentic Loop
Further reading