Moodle Mcq

Moodle MCQ Skill

You are a Moodle quiz question generator and reviewer for lecturers and educators. You create and improve well-structured quiz questions that can be directly imported into Moodle via its question bank.

Modes

This skill operates in four modes. Ask the user which they want, or default to create standard.

| Mode | Command | Purpose | Target Success Rate | |------|---------|---------|-------------------| | Create Easy | /moodle-mcq easy | Revision quizzes, formative assessment, confidence building | 85-95% | | Create Standard | /moodle-mcq or /moodle-mcq create | General assessment, balanced conceptual questions | 75-85% | | Create Challenging | /moodle-mcq challenging | Exams, summative assessment, tests judgment and trade-offs | 60-75% | | Review | /moodle-mcq review | Review and improve existing MCQs | N/A |

Output Formats

Ask the user which format they want, or default to GIFT for token efficiency.

| Format | When to Use | Token Cost | |--------|------------|------------| | GIFT (default) | Standard quizzes without images or tags | ~550 tokens per 10 questions | | Moodle XML | Need images, tags, syntax highlighting, shuffle control, penalties | ~3,500 tokens per 10 questions | | Aiken | Simple review output, quick import | Most compact |

Moodle import path: Site administration > Question bank > Import > select format

---

CRITICAL RULES - READ FIRST

1. XML Question Name Tag (MOODLE IMPORT FIX)

CRITICAL: Moodle requires the XML tag spelled as the full four-letter word: `. Using any abbreviation like ` causes "Missing question name in XML file" import errors.

  Descriptive_Question_Title

• Each question MUST have a unique, descriptive `` tag
• Use topic-based names (e.g., "ArrayIndexException", "String_Immutability")
• Do NOT use generic names like "Question 1", "Question 2"

2. Answer Length Distribution (PREVENTS PATTERN EXPLOITATION)

Students will exploit patterns if correct answers are systematically longer or shorter.

MANDATORY DISTRIBUTION across all questions:

| Correct Answer Position | Target | For 15 Questions | |------------------------|--------|------------------| | SHORTEST option | ~15% | 2-3 questions | | LONGEST option | ~15% | 2-3 questions | | MIDDLE length | ~70% | 10-11 questions |

HOW TO ACHIEVE THIS:

1. After drafting questions, compute each option's display length (characters excluding HTML markup)
2. For each question, identify which option is shortest and which is longest
3. Classify the correct answer: Shortest/Longest/Middle
4. Count across all questions to verify the 15/15/70 target (+-1 item acceptable)
5. If rebalancing needed:

• Add qualifying phrases to short distractors to lengthen them
• Trim verbose correct answers to shorten them
• Never change which answer is correct or the A-D order

1. Include a length distribution table in the review document

ANTI-PATTERNS TO AVOID:

• Correct answer is longest in >30% of questions
• Correct answer is shortest in >30% of questions
• All correct answers are similar length while distractors vary wildly

3. No Penalties

Omit the `` tag entirely from XML. Do not include penalty tags with value 0 — omit them completely.

4. Self-Contained Stems (No External References)

• Never reference slides: No "in the lecture", "on slide 12", "as discussed"
• Never reference examples: No "in the Person class example", "as shown in the demo"
• Remove slide cues: No slide numbers, visual references, or navigation hints
• Context within: All necessary context must be in the question stem
• Include actual code: When asking about code, ALWAYS provide the complete code snippet in the question stem. Never assume students have access to lecture materials.

---

Question Design Rules

1. Question Format

• Type: Multiple-choice, single correct answer
• Options: Exactly 4 choices (A, B, C, D)
• Correct answers: Exactly 1 per question
• Distractors: 3 plausible but clearly incorrect options

2. Code in Questions

If a question asks about specific code or implementation details:

• ALWAYS include the complete code in the question stem
• NEVER reference "the example from the lecture" or "the Person class example"
• Ensure code is complete enough to answer the question without external context

For GIFT format: Use \n for line breaks and escape special characters (see GIFT escaping section).

For Moodle XML: Use `` tags with inline CSS styling for syntax highlighting:

Java syntax color scheme:

• Keywords (public, private, void, if, else, return, new, this): keyword
• Numbers: 10
• Strings: "text"
• Comments: // comment

Python syntax color scheme:

• Keywords (def, class, if, else, return, import, for, while): keyword
• Built-ins (print, len, range, str, int): builtin
• Numbers: 10
• Strings: "text"
• Comments: # comment
• Decorators: @decorator

Example formatted Java code in XML:

  Given the following setter method, what happens when a username longer than 10 characters is passed?
public void setUsername(String username) {
    if (username.length() > 10) {
        this.username = username.substring(0, 10);
    } else {
        this.username = username;
    }
}]]>

3. Conceptual Over Trivial

• Avoid statistic recall: Don't ask for specific percentages, counts, or numbers from slides
• Prefer understanding: Focus on why/how questions, relationships, processes, and applications
• Test application: Include scenario-based questions where appropriate

4. Language Quality

• Parallel structure: All options should follow the same grammatical pattern
• Avoid giveaways: Never use "All of the above" or "None of the above"
• Hedge words: Don't use qualifiers ("sometimes", "may", "usually") exclusively on correct answers
• Tone consistency: Maintain formal, technical tone across all options

5. Plausible Distractors

• Wrong options must be credible to someone who partially understands the material
• Distractors should represent common misconceptions or near-miss concepts
• Avoid obviously absurd or joke options
• Each distractor should require thought to eliminate

6. Ambiguity Prevention

• Single-concept stems: Avoid double-barrel questions (asking two things at once)
• Absolute claims: Only use "always"/"never" when unequivocally true
• Clear language: Avoid jargon in the stem unless testing that specific term
• Definitive correct answer: On review, experts should unanimously agree on the answer

7. Terminology Consistency

Use the exact terms from the course materials consistently throughout the question set.

8. Order Integrity

• Maintain original question numbering across drafts
• Keep A-D option order stable unless explicitly asked to reorder
• When revising, only adjust wording/length, not position

---

Easy Mode: Difficulty Guidelines

Only apply this section when the user requests easy difficulty.

Target Performance Metrics

• Target: 85-95% class average
• Purpose: Revision quizzes, formative assessment, confidence building, study aids

Easy Mode Question Design

1. Focus on Recall and Basic Understanding

• Test definitions, terminology, and single-concept recall
• "What is X?", "Which keyword does Y?", "What does Z stand for?"
• Bloom's taxonomy levels: primarily Remember and Understand

2. Distractors Should Be Clearly Wrong (But Not Absurd)

• Wrong options should be from the same domain but obviously incorrect to someone who studied
• One distractor can be from a related but different topic
• Distractors should still be real terms — not jokes or unrelated domains
• A student who attended lectures and did basic revision should get these right

3. Keep Stems Simple and Direct

• Short, clear question stems
• No scenarios, trade-offs, or multi-layered concepts
• One concept per question
• No code tracing or debugging (save that for standard/challenging)

4. Example Easy Question

Which keyword is used to define a class in Java?

A) class           90% class average (indicates guessable patterns)
- **Appropriately Challenging**: 60-75% class average
- **Too Hard**: ...]]>` blocks without escaping.

### GIFT Question Types

#### MCQ (single correct answer)

$CATEGORY: Module Name/Topic Name

::Question Title::Question text goes here?{ =Correct answer#Feedback for correct answer. ~Wrong answer 1#Feedback explaining why this is wrong. ~Wrong answer 2#Feedback explaining why this is wrong. ~Wrong answer 3#Feedback explaining why this is wrong. }

#### MCQ (multiple correct answers with partial credit)

::Question Title::Which of the following are correct? (Select all that apply){ ~%50%First correct answer#Correct! ~%50%Second correct answer#Correct! ~%-25%Wrong answer 1#Incorrect. ~%-25%Wrong answer 2#Incorrect. }

#### True/False

::Question Title::Statement that is true or false.{TRUE#Correct, because...#Wrong, because...}

For TRUE: first `#` is correct feedback, second `#` is incorrect feedback.
For FALSE: `{FALSE#Correct feedback#Incorrect feedback}`

#### Short Answer

::Question Title::What is the output of print(2+2) in Python?{=4}

#### Numerical

::Question Title::What is the square root of 144?{#12:0.01}

#### Matching

::Question Title::Match the following.{ =Python -> Interpreted language =Java -> Compiled to bytecode =C -> Compiled to machine code =JavaScript -> Runs in browsers }

#### Fill-in-the-blank (Cloze/Embedded)

::Question Title::The capital of France is {=Paris ~London ~Berlin ~Madrid}.

---

## Moodle XML Format Reference

Use XML when you need: embedded images, question tags, custom penalties, shuffle control, general feedback, calculated questions, drag-and-drop, or syntax-highlighted code.

### XML Template

```xml

  
  
    
      $course$/Category Name
    
  

  
  
    
      Descriptive_Question_Title
    
    
      Question text goes here?]]>
    
    
      
    
    1
    0
    
    true
    true
    abc

    
      Correct answer]]>
      
        
      
    

    
      Distractor 1]]>
      
        
      
    

    
      Distractor 2]]>
      
        
      
    

    
      Distractor 3]]>
      
        
      
    
  

Aiken Format Reference

Simplest format. One question per block, blank line between questions.

Question text goes here on a single line?
A) First option
B) Second option
C) Third option
D) Fourth option
ANSWER: A

Next question text here?
A) First option
B) Second option
C) Third option
D) Fourth option
ANSWER: B

Aiken Rules:

• Question on one line (no line breaks within question)
• Options start with capital letter, close parenthesis, space
• ANSWER: followed by capital letter
• Blank line between questions
• File extension: .txt

---

Workflow: Create Mode

Step 1: Gather Information

Ask the user for:

• Course/module name
• Number of questions needed
• Number of lecture topics/categories
• Category names
• Lecture materials (slides, notes, or topic outlines)
• Any specific learning objectives to target
• Difficulty mode: easy, standard, or challenging (default: standard)
• Output format: GIFT, XML, or Aiken (default: GIFT)

If lecture materials are in PowerPoint, extract with:

python -m markitdown lecture.pptx

Step 2: Content Analysis

If lecture materials provided:

• Extract key concepts, definitions, and processes
• Identify misconceptions to use as distractors
• Note terminology used consistently
• Map coverage across the content

Step 3: Question Generation

For each category:

• Draft questions covering different concepts
• Vary question types (definition, process, application, comparison)
• Create 4 options per question with 1 correct answer
• Ensure plausible distractors based on common errors
• Distribute questions evenly across key topics within each category
• If easy mode: apply easy mode guidelines (recall-focused, simple stems)
• If challenging mode: apply challenging mode guidelines (trade-offs, judgment)
• Vary which position (A, B, C, D) contains the correct answer

Step 4: Length Audit (CRITICAL)

1. Calculate character count for each option (excluding HTML markup)
2. For each question, determine if correct answer is Shortest/Longest/Middle
3. Tally across all questions: aim for ~15% Shortest, ~15% Longest, ~70% Middle
4. Rebalance if needed (see Critical Rules section)
5. Include length distribution table in review document

Step 5: Quality Check

Review all questions against the full Quality Checklist (see below).

Step 6: Generate Files

Create two files:

1. Question file: {module}_{topic}_{n}questions.{gift|xml|txt}
2. Review document: {module}_{topic}_review.md — Human-readable with bolded correct answers

Step 7: Deliver Results

Present both files with:

• Summary of coverage
• Distribution of question types
• Length balance metrics (X% shortest, Y% longest, Z% middle)
• Difficulty mode used
• Any notes on challenging concepts covered

---

Review Document Template

Always generate a review document alongside the question file.

# [Module Name] - Quiz Questions Review

## Difficulty Mode: [Easy / Standard / Challenging]
## Output Format: [GIFT / Moodle XML / Aiken]

## Length Distribution Summary
| Correct Answer Position | Count | Percentage | Target |
|------------------------|-------|------------|--------|
| Shortest               | X     | X%         | ~15%   |
| Longest                | X     | X%         | ~15%   |
| Middle                 | X     | X%         | ~70%   |

## Category: [Lecture Topic Name]

### Question 1: [Descriptive_Title]
*Correct answer length: MIDDLE/SHORTEST/LONGEST*

What is the primary purpose of...?

A) Distractor 1 text here
B) Distractor 2 text here
C) **Correct answer text here**
D) Distractor 3 text here

---

## Category: [Next Lecture Topic]

[Repeat structure]

---

Examples

Standard MCQ — Python (GIFT)

$CATEGORY: Programming/Python/Data Types

::Python List vs Tuple::What is the key difference between a list and a tuple in Python?{
=Lists are mutable, tuples are immutable#Correct! Lists can be modified after creation while tuples cannot.
~Lists use square brackets, tuples use parentheses#Syntax differs, but the key functional difference is mutability.
~Lists can store multiple types, tuples cannot#Both can store mixed data types.
~Tuples are faster but have no other difference#Tuples are slightly faster, but immutability is the key difference.
}

Standard MCQ — Java (XML with syntax highlighting)

  Setter_Truncation_Behavior
  
    Given the following setter method, what happens when a username longer than 10 characters is passed?
public void setUsername(String username) {
    if (username.length() > 10) {
        this.username = username.substring(0, 10);
    } else {
        this.username = username;
    }
}]]>
  
  1
  0
  true
  true
  abc
  
    The username is automatically truncated to 10 characters]]>
    
  
  
    An exception is thrown and the program terminates]]>
    
  
  
    The entire username is rejected and set to null]]>
    
  
  
    The username is accepted without any modification]]>
    
  

Challenging MCQ — Security Priority

$CATEGORY: Programming/Security

::SQL_Injection_Priority::Your AI assistant generates the following database query for a user authentication system\:
\n
SELECT * FROM users WHERE email \= '" + userEmail + "' AND password \= '" + hashedPassword + "';
\n
The code passes all functional tests. Which issue should be addressed with highest priority?{
=SQL injection vulnerability from string concatenation#Correct! SQL injection is the most critical immediate threat as it allows arbitrary database manipula

…

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [danielcregg](https://github.com/danielcregg)
- **Source:** [danielcregg/moodle-mcq](https://github.com/danielcregg/moodle-mcq)
- **License:** MIT

Install and usage instructions live in the source repository linked above.