Courses that test the learner. Every chapter ends with a 10-question quiz that grades itself in the page, by level, difficulty and topic. Paste the results back and Claude names the misconceptions, says what to re-read, and builds a retest aimed at the weak spots.
$ npx skills add akshaypjoshi/course-studio -g -a claude-code $ npx skills add akshaypjoshi/course-lab -g -a claude-code
/plugin marketplace add akshaypjoshi/course-lab /plugin install course-studio@course-lab /plugin install course-lab@course-lab
This is the quiz from the end of Course Studio's demo chapter on hashing, running live. Answer the ten questions, tick I'm guessing when you are, and press Submit. You get a level, a breakdown by difficulty and topic, and a link to the section to re-read for every weak topic.
lab.py from one quiz.jsonSee it at the end of the chapterCourse Studio writes the course; Course Lab adds the quizzes, the grading and the follow-up.
"The binary search chapter, with a quiz"
The full chapter with its figure, then 10 questions written from the finished text, verified, and added as its last section.
"A short course on rate limiting, with tests"
A course plan, its chapters, a quiz for each, and a final test across them once two chapters are written.
"Add a quiz to chapter 5"
For a chapter that already exists. Every question links back to the section that teaches it.
"course-lab results ..."
Paste the copied results. Misconceptions first, sections to re-read, concrete practice, and a new 10-question retest page.
The quiz tells the learner where they stand right away; the evaluation tells them why.
Course Studio writes and checks it: run code, an interactive figure, a built page.
3 easy, 4 medium, 3 hard from the finished chapter. Every wrong option is a named mistake with a reason, and every answer is computed or run.
Weighted score and level, by difficulty and by topic, confident mistakes apart from guesses, and a Revisit link for each weak topic.
Claude reads which wrong options were picked, names the misconception behind them, and builds a retest weighted to the weak topics.
Easy, medium and hard are worth 1, 2 and 3 points, so the level reflects depth. Mastered needs 85% and at least two of the three hard questions.
course-lab results Quiz: Hashing, chapter test [dsa-05-hashing] · Data Structures & Algorithms Score: 4/10 · weighted 6/20 (30%) · level: Needs review Difficulty: easy 2/3, medium 2/4, hard 0/3 Answers: Q1 [easy · Load factor & resizing] picked C: correct (guessed) Q2 [easy · Hash functions] blank: wrong, answer B Q6 [medium · Pitfalls] picked A: wrong, answer B (was sure) ...
Course Studio's Hashing chapter with Course Lab's quiz at the end, rebuilt by both skills' own scripts on every deploy.
Load factor, chaining, open addressing with tombstones, resizing and the inputs that break a hash map, then ten questions from a one-line load factor to a naive delete that breaks linear probing.
Course Lab writes the quizzes and uses Course Studio for everything else, so install both. It finds Course Studio by itself.
| Tool | Needed for |
|---|---|
| Course Studio | Planning, writing, figures and page builds |
Python 3.9+ with markdown, pyyaml, pygments | Building pages (pip install -r requirements.txt in the skill folder) |
playwright and Chromium | Taking every quiz headless in the check |
| Node.js 20.10+ | Course Studio's figures |