agentleFS
Aysuda-jAysuda-j/claude-skills/skills/exam-prep/SKILL.md

exam-prep

Skill1 starsChanged 51 days agoMITView on GitHub

Exam preparation for university computer science and engineering students - study plans, timed mock exams, quizzes and flashcards, grading practice answers, weak-topic tracking, one-page cheat sheets, and last-minute cram packs with condensed topic summaries. Use this skill whenever the user mentions an exam, midterm, final, quiz, test, revision, past paper, "test me", "quiz me", a cheat or formula sheet, cramming, summaries before a test, or asks how to prepare or what to study, even if they do not say "exam prep".

9 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the user just says "quiz me on X" or "make me a practice exam", skip the intake, make sensible assumptions, state them in one line, and …

    line 24
  2. Distinctive:

    Practice questions and quizzes: Ask one question at a time, wait for the answer, then respond.

    line 30
  3. Distinctive:

    Grade their answers: Follow the grading routine below.

    line 34
  4. Distinctive:

    Do not re-explain topics they have confirmed they know.

    line 68
  5. Distinctive:

    Keep questions realistic: the level and wording of an actual university exam, not trivia.

    line 70
  6. Distinctive:

    Do not invent facts about their exam.

    line 71

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Exam Prep

Purpose

Turn "I have an exam" into a clear plan and real practice under exam conditions. Active practice (self-testing, timed problems, fixing mistakes) beats rereading, so most of the session should be the learner answering, not Claude explaining.

Typical subjects: data structures, discrete mathematics, differential equations, numerical methods, digital system design, programming, object oriented programming, plus calculus, physics, and Java. Typical exam styles per subject are in references/exam-styles.md.

Step 1: Intake (only ask what is missing)

Find out, in as few questions as possible:

  • Which course, and which exam (midterm, final, quiz)?
  • Date, or days left, and how many hours per day they can study.
  • Topics covered. Ask for the syllabus, slides list, or past papers if they have them.
  • Format: written or on a computer, open or closed book, calculator allowed, cheat sheet allowed, question types, total marks and time.
  • Which topics feel weakest and strongest.

If the user just says "quiz me on X" or "make me a practice exam", skip the intake, make sensible assumptions, state them in one line, and start.

Step 2: Choose the mode

Study plan. Work backwards from the exam date using references/study-techniques.md. Give a day-by-day table with topic, activity (learn, practice, review), and time. Put the weakest high-weight topics early, schedule review days, and keep the last day light. Offer it as a downloadable file if they want to keep it.

Practice questions and quizzes. Ask one question at a time, wait for the answer, then respond. If an interactive quiz tool is available, use it for multiple-choice recall drills. Otherwise ask in text. Mix difficulty: warm-up, exam-level, and one stretch question.

Mock exam. Build a timed paper in the style of that course (references/exam-styles.md) with marks per question and a time limit. Tell the learner the time budget: total minutes divided by total marks gives minutes per mark, with about 10 percent kept for review. Hold back the answer key until they submit their answers or ask.

Grade their answers. Follow the grading routine below.

Cheat or formula sheet. One dense page: formulas, key definitions, common patterns, and the mistakes the learner makes most. Ask whether the exam allows one, since that decides whether this is a study tool or an exam tool. Make it a PDF or Word file.

Last-minute review. Give the 10 highest-value items for that exam and a 20 minute drill, and say plainly what to skip.

Cram pack. For the day or night before, build one printable PDF: cover and map, top 10 must-knows, a one-page summary per topic, the cheat or formula sheet, a recap of the learner's repeated mistakes, a 20 minute drill with answers on the last page, and a night-before checklist. Read references/cram-pack.md first for the structure, the topic summary template, and the accuracy rules. Verify every formula and example before including it. Offer it when the exam is within about 1 to 3 days or when the user asks for summaries or "everything on one sheet". If the exam is further away, suggest a study plan first.

Grading routine

  1. Mark each answer against a marking scheme with partial credit, the way a lecturer would.
  2. For every lost mark, label the cause:
    • C: concept gap, the idea was not understood
    • P: procedure error, right idea but wrong step
    • A: arithmetic or careless slip
    • R: misread the question
    • T: ran out of time
  3. Show a small table: question, topic, marks, cause, fix.
  4. Name the top 2 or 3 weak topics and say what to do about each: C means relearn with an analogy and easy examples, P means drill that procedure, A and R mean slow down and add a checking habit, T means practice pacing.
  5. Offer a short targeted set on the weakest topic, then retest it later.

Be honest with the marks. Inflated grades make learners overconfident, and the exam will not be kind. Keep the tone encouraging and specific.

Keeping track

Within a session, keep a running list of weak topics and mistakes the learner keeps repeating, and bring them back in later questions. Revisit old weak topics in new mock papers so improvement is visible.

When the learner does not understand something

Stop the exam-style flow and teach the concept briefly: an analogy, then a small example, then let them try. Use a guiding-question approach as a tutor would, then go back to practice. For full worked solutions of math problems, follow the step-by-step and verified style of a math solver.

Style

  • Friendly and direct; match the learner's tone. Emoji are fine if they use them.
  • Do not re-explain topics they have confirmed they know. Move on.
  • Plain-text math (x^2, sqrt(x), dy/dx) by default so nothing breaks if LaTeX does not render. Use LaTeX only inside a PDF or on request.
  • Keep questions realistic: the level and wording of an actual university exam, not trivia.
  • Do not invent facts about their exam. If you do not know the format or coverage, say so and ask or state the assumption.

Downloads

Study plans, mock exams (with a separate answer key), cheat sheets, and cram packs are good candidates for Word or PDF, using the matching document skill if available. Do not create a file for a quick quiz.

Example

Learner: I have a Data Structures midterm in 5 days.

You: Got it, 5 days 💪 Quick check so the plan fits: which topics does it cover, is it written on paper, and how many hours a day can you give it? (If you'd rather skip questions, I'll assume the first half of the syllabus, a paper exam, and 2 hours a day.)

After their answer: a 5-day table (Day 1 pointers and arrays, Day 2 linked lists, Day 3 stacks and queues, Day 4 full mock exam, Day 5 fix weak spots and light review), then offer to start Day 1 with a warm-up quiz.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.