Lesson QuizSAFE

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

graph TD
A["Select lesson"] --> B["Choose timing: pre/during/after"]
B --> C["Answer 10 questions in 5 rounds"]
C --> D["Score and review feedback"]
style A fill:#4CAF50,color:#fff
style D fill:#2196F3,color:#fff

/lesson-quiz [lesson-name-or-number]

/lesson-quiz hooks
/lesson-quiz 03
/lesson-quiz advanced-features
/lesson-quiz           # (prompts for lesson selection)

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: lesson-quiz
description: "Test a learner on a single Claude Code tutorial lesson (01-10) with 10 questions, scoring answers and flagging weak spots. Use before, during, or after a lesson. Don't use for whole-tutorial assessment or explaining a topic instead of testing it."
effort: high
metadata:
  version: 1.4.0
  author: Luong NGUYEN
---

# Lesson Quiz

Interactive quiz that tests understanding of a specific Claude Code lesson with 8-10 questions, provides per-question feedback, and identifies areas to review.

## Prerequisites

This skill requires:

- The tutorial repo checked out, so the lesson directories `01-slash-commands/` ... `10-cli/` and each `README.md` are readable.
- `references/question-bank.md` present in this skill (the source of all questions).

Before starting, confirm the target lesson's `README.md` exists. If it is missing, do not fabricate questions — warn the user and ask them to check the repository structure (see Error Handling).

**Guardrails:**

- **Never invent questions or answers.** Ask only questions drawn from `references/question-bank.md` for the selected lesson; if the bank lacks entries for a lesson, say so rather than making them up.
- **Score accurately.** Track which shuffled position holds the correct answer for every question and validate each response against it — never guess a score.
- **Confirm timing before scoring.** The pre/during/after choice changes how results are framed; do not skip it.

## Instructions

### Step 1: Determine the Lesson

If the user provided a lesson as an argument (e.g., `/lesson-quiz hooks` or `/lesson-quiz 03`), map it to the lesson directory:

**Lesson mapping:**
- `01`, `slash-commands`, `commands` → 01-slash-commands
- `02`, `memory` → 02-memory
- `03`, `skills` → 03-skills
- `04`, `subagents`, `agents` → 04-subagents
- `05`, `mcp` → 05-mcp
- `06`, `hooks` → 06-hooks
- `07`, `plugins` → 07-plugins
- `08`, `checkpoints`, `checkpoint` → 08-checkpoints
- `09`, `advanced`, `advanced-features` → 09-advanced-features
- `10`, `cli` → 10-cli

If no argument was provided, present a selection prompt using AskUserQuestion:

**Question 1** (header: "Lesson"):
"Which lesson do you want to quiz on?"
Options:
1. "Slash Commands (01)" — Custom commands, skills, frontmatter, arguments
2. "Memory (02)" — CLAUDE.md, memory hierarchy, rules, auto memory
3. "Skills (03)" — Progressive disclosure, auto-invocation, SKILL.md
4. "Subagents (04)" — Task delegation, agent config, isolation

**Question 2** (header: "Lesson"):
"Which lesson do you want to quiz on? (continued)"
Options:
1. "MCP (05)" — External integration, transport, servers, tool search
2. "Hooks (06)" — Event automation, PreToolUse, exit codes, JSON I/O
3. "Plugins (07)" — Bundled solutions, marketplace, plugin.json
4. "More lessons..." — Checkpoints, Advanced Features, CLI

If "More lessons..." is selected, present:

**Question 3** (header: "Lesson"):
"Select your lesson:"
Options:
1. "Checkpoints (08)" — Rewind, restore, safe experimentation
2. "Advanced Features (09)" — Planning, permissions, print mode, thinking
3. "CLI Reference (10)" — Flags, output formats, scripting, piping

### Step 2: Read the Lesson Content

Read the lesson README.md file to refresh context:
- Read file: `<lesson-directory>/README.md`

Then use the question bank from `references/question-bank.md` for that lesson (10 pre-written questions with answers and explanations). To stay within the context budget, read only the one selected lesson's README — not all ten.

### Step 3: Present the Quiz

Ask the user about quiz timing context:

Use AskUserQuestion (header: "Timing"):
"When are you taking this quiz relative to the lesson?"
Options:
1. "Before (pre-test)" — I haven't read the lesson yet, testing my prior knowledge
2. "During (progress check)" — I'm partway through the lesson
3. "After (mastery check)" — I've completed the lesson and want to verify understanding

This context affects how the results are framed (see Step 5).

### Step 4: Present Questions in Rounds

Present 10 questions from the question bank in rounds of 2 questions each (5 rounds total). Each question uses AskUserQuestion with the question text and 3-4 answer options.

**IMPORTANT**: Use AskUserQuestion with max 4 options per question, 2 questions per round.

For each round, present 2 questions. After the user answers each round, immediately show per-question feedback: whether each answer was correct or incorrect, and if incorrect, show the correct answer and a brief explanation. Then proceed to the next round. After all 5 rounds, proceed to final scoring.

**Question format per round:**

Each question from the question bank has:
- `question`: The question text
- `options`: 3-4 answer choices (one correct, labeled in the bank)
- `correct`: The correct answer label
- `explanation`: Why the answer is correct
- `category`: "conceptual" or "practical"

**CRITICAL — Shuffle answer options**: For each question, you MUST randomize the order of the answer options before presenting them via AskUserQuestion. Do NOT present them in the order they appear in the question bank (A, B, C, D), and do NOT place the correct answer first. Use a different random permutation for each question. Track which shuffled position contains the correct answer so you can score accurately.

Example: If the question bank lists options A (correct), B, C, D — you might present them as: C, A, D, B. The correct answer is now in position 2.

Present each question using AskUserQuestion. Record the user's answer for each.

### Step 5: Score and Present Results

After all rounds, calculate the score and present results.

**Scoring:**
- Each correct answer = 1 point
- Total possible = 10 points

**Grade scale:**
- 9-10: Mastered — Excellent understanding
- 7-8: Proficient — Good grasp, minor gaps
- 5-6: Developing — Fundamentals understood, needs review
- 3-4: Beginning — Significant gaps, review recommended
- 0-2: Not yet — Start from the beginning of this lesson

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.