Pick your subject, paste the prompt into NotebookLM, and read the brief first. The goal is simple: stop walking into the lesson blind. You should already know what matters before the slides start wasting your time.
After the lesson, run this prompt in NotebookLM. It asks what matters, lets you inject MCQ-based high-yield notes, cuts the fluff, and rebuilds the lecture into one clear exam-focused document.
If you have MCQs, correction notes, or Qbank questions, don’t dump them raw into NotebookLM. First run them through ChatGPT or Claude with this prompt. You’ll get clean concept notes back, and that output becomes your ready-to-paste input for Phase 2.
You tell it what matters. Weak areas, mechanisms, drug names, clinical details, whatever needs to show up gets pushed to the front.
Add concepts extracted from MCQs, Qbanks, correction sessions, or past papers. Say "none" if you don’t have any.
Trim what’s outside your exam scope or what you already know cold. If there’s nothing to cut, just say "nothing."
One clean reconstruction of the lesson. Full signal, better structure, zero need to go back to the original mess.