Verified against ChatGPT · 2026-08-08
Turn a stack of graded past papers into a revision plan that spends your remaining days on what's actually costing you marks
Feeds in your topic-by-topic scores from past papers or mock tests and gets back a day-by-day revision schedule weighted toward your weakest, highest-value topics instead of a generic full-syllabus re-read.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are building me a revision schedule for the exact number of days I have left before my exam, based on where I'm actually losing marks, not a generic re-read of the whole syllabus in order. EXAM AND DATE CBSE Class 12 Physics board exam, 3 March 2027 DAYS AVAILABLE AND DAILY STUDY HOURS 18 days left, roughly 3 usable hours on weekdays and 6 on weekends TOPIC-BY-TOPIC PERFORMANCE Electrostatics 40%, Current Electricity 55%, Optics 85%, Modern Physics 35%, EM Induction 60% MARK WEIGHTING OF EACH TOPIC ON THE ACTUAL EXAM Electrostatics ~16% of paper, Current Electricity ~10%, Optics ~14%, Modern Physics ~10%, EM Induction ~12% NON-NEGOTIABLE COMMITMENTS DURING THIS WINDOW Two other board exams (Chemistry, Math) in the same window needing roughly 1 hour/day each RULES Rank topics by a combination of how many marks I'm losing on them and how many marks they're worth on the real exam — a topic I'm weak on but that's worth 2% of the paper should get less time than a topic I'm weak on that's worth 15%, even though both are "weak." Do not default to allocating time evenly across topics; state explicitly which topics are getting deliberately less time than their syllabus weight would suggest, and say why. Build in at least one full practice-paper sitting under timed conditions before the last two days, and reserve the final day for review only, not new material. For each day, name the specific topic, what "done" looks like for that day (e.g. "can solve all past-paper questions on this topic from the last 3 years without looking at notes"), and roughly how long it should take, adjusted to the hours I actually have that day. If the days available clearly aren't enough to adequately cover every weak topic at the depth the mark weighting would ideally deserve, say so plainly and tell me which topics you're deliberately triaging down to a lighter pass or dropping, rather than quietly producing an overstuffed schedule that looks complete but isn't realistic to finish. OUTPUT FORMAT A short paragraph stating the triage logic and which topics got cut or compressed and why, followed by a day-by-day table: Day | Topic | Goal for the day | Time budget | Resource type (notes review / practice questions / timed paper).
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
GPT-5.1 defaults to a well-meaning but structurally lazy move when asked for a "revision plan": it distributes days roughly evenly across whatever topic list it's given, because even coverage is the safest-looking answer and requires no judgment call about what to cut. Forcing it to combine two numbers per topic — weakness and mark-weighting — rather than one breaks that default, because there's no longer a single sortable list to spread evenly across; it has to actually multiply out which combination of low-score-and-high-weight topics deserves disproportionate time, which is a genuine reasoning step rather than a formatting one. The explicit instruction to name what's being triaged down or dropped matters because an unconstrained plan that's asked to cover everything in too few days will silently pad each topic's allotted time until the total adds up to the available hours, producing a schedule that looks complete on paper but was never checked against whether a real day's block is achievable — naming the triage forces the model to do that arithmetic check out loud instead of eliding it. The "done" criterion per day matters mechanically too: without a concrete exit condition, GPT-5.1's day-by-day plans tend to describe activities ("review electrostatics") rather than checkable outcomes, which means the plan can't tell you whether you're actually on schedule three days in, only whether you followed the itinerary.
What you get back
Triage note: Modern Physics (35% score, only 10% of paper) and Electrostatics (40% score, 16% of paper) get priority; Optics (85% score) drops to a single 90-minute refresh on Day 14 instead of a full day, since further gains there are marginal. Day 1 — Electrostatics: solve every past-paper question from 2022-2025 unaided, ~3 hrs. Day 2 — Electrostatics (cont.): redo only the ones missed yesterday plus Gauss's law edge cases, ~2 hrs...
Verified against
ChatGPT GPT-5.1 · 2026-08-08
Changelog
- 2026-08-08 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
