Generate the spacing calendar, not another worksheet
The highest-value revision artefact isn't more questions. It's the schedule that says which old topic resurfaces when.
Aug 19, 2026 · 1 min read
Distributed practice is one of the best-evidenced effects in learning science: a meta-analysis of 254 studies shows spaced review beats massed review, with the optimal gap scaling with how long the material must last (Cepeda et al., 2006; rated a top-utility technique by Dunlosky et al., 2013). The hard part was never believing it; it was the bookkeeping. That's now a prompt. (Manning's spaced-practice-scheduler skill automates exactly this.)
In practice
Here are my topics for the term and my 4 lessons per week:
[topics]. Exam on [date]. Build a week-by-week grid where
each lesson opens with 3 retrieval questions from earlier
topics, gaps expanding from 1 week to 3, every topic
resurfacing at least 3 times before the exam.- Feed the output straight into your do-now starters (see "Have a task on the board before they sit down"). The calendar decides the questions; you stop improvising them.
- Sanity-check the gaps: early reviews close together, later ones further apart, nothing seen once and never again.
Quick takeaway: let AI do the bookkeeping of forgetting so your starters can fight it.