All products

Plan exam revision with AI

Organise authorised course material into sourced topics, practice priorities and a short revision plan, with uncertainty attached to every prediction.

Plan exam revision with AI around the materials your course actually provides. The finals-week skill prepares notes, assignments and released papers, then maps topics, formulas and question patterns back to their sources. PDF and image preparation uses a documented local MinerU service.

The resulting plan can include drills and text flashcards. Repeated coverage is evidence for prioritisation, not a guarantee of what an examiner will ask. The workflow supports preparation, excluding live-exam assistance and unreleased questions.

What it does

  • Organise notes and images into a course workspace
  • Map definitions and formulas to their source
  • Rank revision priorities with supporting evidence
  • Choose a short plan for the remaining time
  • Separate essential drills from optional practice
  • Prepare text or TSV revision cards

Install and use

  1. Provide the complete package, Python3.10+, requests and a compatible running local MinerU API. The documented legacy contract is supported by version3.1.15; newer APIs differ. MinerU and its models are separately acquired under their own terms.
  2. Try: “Use these notes and released papers for a three-day revision plan. Attach source locations and confidence to every predicted priority.”
  3. Inspect extracted text, especially formulas and scanned pages. Recognition and language support depend on the configured service; unreadable material must not become a confident claim.
  4. Review the plan, drills and cards in the output workspace. The default recipient is local; configuring another API address changes where files go. Resolve important extraction uncertainty before relying on the agent’s analysis.

FAQ

Source

Made by Rain-dust. Original skill: finals-week, published on GitHub under the MIT licence.

View on GitHub