Interactive vs. non-interactive development

Two fundamentally different modes of running a harness. In interactive mode the
human types a prompt, reviews the output, and continues the loop — staying in
control. In non-interactive (headless) mode the terminal UI is stripped away and
the agent runs autonomously, with no prompt waiting for a response — the basis for
running agents on schedules, inside CI/CD pipelines, and as background services.
Students learn the human-in-the-loop versus human-on-the-loop distinction and the
prevailing hybrid default: plan interactively, execute autonomously in a sandbox,
verify in CI, and checkpoint at PR review.