Build a study plan that targets what you'll actually forget

Stop re-reading everything equally and spend your time where you're actually going to lose it.

Use it for: Anyone studying for an exam, certification, or just trying to actually retain what they read.

The default study plan is "read the material once, then review it before the exam" — which treats every topic as equally likely to be forgotten and equally important, when neither is true. Some things you understand deeply the first time and barely need to touch again; others slip within days no matter how well you understood them in the moment, because they're arbitrary, counterintuitive, or disconnected from anything you already know. A study plan that doesn't distinguish between these wastes your best resource — time before the exam — on the wrong material.

This prompt asks the model to make that distinction explicit: which topics are forgetting-prone, and why, so the schedule can put spaced review where it actually pays off instead of everywhere equally.

When not to use this

Needs a real topic list, not "help me study for my exam" — the more specific your material, the more specific and useful the forgetting predictions. And it's a planning tool, not a tutor: pair it with something like a topic-explanation prompt for the material itself.

Fill in

{{exam}} — What you're studying for{{date}} — When it is{{topics}} — What you need to cover{{hours_per_week}} — Roughly how many hours a week you have

Prompt

I'm studying for {{exam}}, happening on {{date}}. Here's what I need to
cover: {{topics}}

I have roughly {{hours_per_week}} hours a week until then.

Build me a study plan, but don't split time evenly across topics. For
each topic:

1. Rate how forgetting-prone it likely is — arbitrary facts and
   disconnected details forget faster than concepts that follow logically
   from something else. Give a one-line reason for each rating.
2. Schedule initial coverage, then spaced review sessions timed to hit
   each topic again right before you'd naturally start forgetting it —
   sooner for high-forgetting topics, later and less often for topics
   that stick once understood.
3. Flag topics that depend on another topic being solid first, and
   sequence accordingly.
4. Build in one full practice run-through in the final week, and tell me
   honestly whether my available hours are enough to cover everything
   properly — if not, tell me what to cut or triage rather than
   silently thinning every topic to fit.

Give me the plan as a week-by-week schedule I can actually follow, not
just a table of ratings.

Why it works

This is spaced repetition applied deliberately instead of by accident. Most self-study review is either "read it again the night before" (too late to fix real gaps) or "review everything on the same schedule" (wastes time on what already stuck). Timing review sessions to land right as forgetting would naturally kick in is the mechanism spaced repetition systems are built around; doing it explicitly by topic, rather than one blanket schedule, is what most self-made study plans skip.

The honest-capacity check at the end matters more than it looks. A plan that pretends your seven hours a week can properly cover forty hours of material isn't a plan, it's a wish — and finding that out in week one, when you can still triage, is far better than finding it out the week before the exam.

Common mistakes

×Listing topics too broadly ("networking") instead of specifically enough for the forgetting ratings to mean anything ("VPC peering vs. security group rules").
×Ignoring the capacity-mismatch warning and trying to cover everything anyway. If the plan says triage, triage — a shallow pass on everything beats a real pass on eighty percent that ends in panic.
×Skipping the final practice run-through because the schedule feels tight. It's the step that tells you whether the studying actually worked, and it's cheaper to find gaps there than in the exam room.

Follow-ups

I'm behind on week 2 — replan the rest, prioritizing the highest-forgetting topics.
Turn the highest-forgetting topics into flashcard-style questions I can drill.

Did this work?

You can articulate why each topic is scheduled when it is, and the plan puts more time on the things you're actually shaky on — not equal time on everything, front-to-back.

Tested on claude-opus-5. Evidence status is draft; it moves to battle-tested only on recorded runs, never by hand.

Our sponsor

Consense — Big decision? Compare and combine multiple LLM responses.