What Needs Grading, and What Doesn't
Most Step types need nothing from you. Deliver steps complete on their own. Validate steps mark themselves against the answers you set. Collect steps gather responses without a right or wrong. The place a Game Master's actual review time goes is Open Text steps and the written submissions they produce.
Setting the Bar Before They Submit
An Open Text step can require a minimum word count, so a one-line answer simply can't be sent. You never see the half-effort version, and the player gets the feedback immediately rather than waiting for you to bounce it back.
Turning On AI-Assisted Feedback
If your Adventure has an AI API key configured, an Open Text step can be validated automatically against criteria you write in plain language — "the answer must name at least two causes of the French Revolution", say. The player gets that feedback and can revise before submitting to you. It doesn't replace your judgment on the final grade; it means what reaches you has usually already had one round of self-correction.
You can also set how strictly the AI reads the criteria, which is worth tuning on a real cohort's answers rather than guessing up front.
Your Review Queue
Submissions land against the Milestone, carrying the player's text and, if you're using graded Milestones, a grade you assign — completion, percentage, or letter, matching whatever grading mode you set when you designed it. Work through submissions per-Milestone rather than per-player if you're grading a cohort — it's easier to stay consistent when you're comparing several answers to the same prompt back to back.
Where to go next: Reviewing & Grading Submissions in Bulk, once you're grading at cohort scale.