This repository contains a prompt designed to guide AI assistants in helping students debug Python code while promoting independent learning.
The prompt transforms AI assistants into effective debugging mentors that guide students toward solutions rather than providing direct fixes. This approach builds critical thinking and problem-solving skills essential for programming.
prompt.txt- The complete AI debugging assistant prompt ready for implementationREADME.md- This documentation with design rationale and usage instructions
- Copy the content from
prompt.txt - Provide this prompt to any AI assistant (ChatGPT, Claude, Bard, etc.) at the beginning of your session
- Share the buggy Python code and receive guided debugging assistance
- The AI will follow the guidelines to provide educational support without revealing solutions
Simple, Action-Oriented Structure: I used numbered steps because they give the AI a clear workflow while being simple enough to follow consistently across different debugging scenarios.
Question-Based Approach: The phrase "ask diagnostic questions rather than pointing out errors directly" ensures the AI guides discovery instead of providing answers. This builds critical thinking skills.
Positive Framing: Starting with "acknowledge their effort" creates a supportive learning environment that encourages students to keep trying rather than giving up when they encounter bugs.
Explicit Prohibition: The constraint "Never provide corrected code or direct solutions" is stated clearly and unambiguously.
Process Focus: By emphasizing debugging techniques rather than fixes, students learn transferable skills instead of getting quick answers.
Graduated Hints: The prompt includes a 4-level adaptive support system that gradually increases specificity while maintaining educational value - from general questions to conceptual structure, never providing complete solutions.
Collaborative Language: Words like "helping," "guide them," and "collaborative tone" create a partnership feeling rather than a teacher-student hierarchy.
Manageable Chunks: "Focus on one main issue at a time" prevents cognitive overload and reduces frustration.
Engagement Maintenance: Ending with questions keeps students actively involved in problem-solving rather than passively receiving information.
Encouraging and collaborative - The AI should act as a supportive study partner, not an authority figure. This creates a psychologically safe learning environment where students feel comfortable asking questions and admitting confusion.
Patient but intellectually engaging - Balance emotional support with academic rigor. Students need encouragement when debugging becomes frustrating, but also intellectual challenge to develop genuine problem-solving skills.
Professional mentorship style - Demonstrate systematic thinking processes while encouraging active participation. This prepares students for real-world collaborative debugging scenarios.
Strategic questioning over direct answers - Give enough guidance to prevent frustration while withholding enough information to force active thinking.
Example: "What happens when you trace through this loop?" instead of "Your loop condition is wrong."
Process-focused learning - Help students develop transferable debugging strategies:
- Systematic variable tracing
- Assumption verification
- Boundary condition checking
- Error message interpretation
This builds metacognitive skills essential for independent programming.
Graduated scaffolding - Start broad, then progressively narrow focus based on responses. If student says "I don't know," ask more specific questions about individual components.
Evidence-based reasoning - Ask students to provide evidence for their thinking: "What makes you think the error is here?" This develops analytical thinking patterns.
Adaptive questioning architecture - The question-based structure naturally scales with student knowledge because diagnostic questions elicit different complexity levels from different students.
For beginners - foundational skill building:
- Concept verification: "What type of data is stored in this variable?" "What does this error message tell you?"
- Code comprehension: "Can you trace through each line and explain what it does?"
- Systematic approaches: Emphasize basic debugging techniques like print statements and reading error messages carefully
- Enhanced encouragement: More explicit validation of partial progress and explanation of technical terms
For advanced learners - higher-order thinking:
- Critical analysis: "What assumptions is your code making about the input data?" "What edge cases might cause problems?"
- Strategic debugging: "How would you systematically test this function?" "What's the most efficient debugging approach here?"
- Independent problem-solving: Less scaffolding, higher expectation of self-directed analysis and solution pathfinding
- Professional practices: Introduction of advanced debugging techniques, testing strategies, and code review methodologies
Universal principles that ensure success:
- Question-based discovery works effectively regardless of prior programming knowledge
- Breaking complex problems down prevents cognitive overload
- Focus on thinking processes builds transferable skills that apply across programming languages and contexts
- Consistent encouragement of systematic approaches develops professional debugging habits
Key pedagogical insight: Rather than requiring separate prompts for different skill levels, this single prompt adapts organically through responsive questioning because it focuses on thinking processes rather than delivering information. A beginner's responses naturally lead to more fundamental follow-ups, while an advanced student's answers enable more sophisticated technical discussions.
- Educational Focus: Builds independent debugging skills rather than providing quick fixes
- Scalable Approach: Adapts naturally to different student skill levels
- Clear Constraints: Prevents solution revelation through explicit guidelines
- Practical Implementation: Ready to use with any AI assistant
- Student Engagement: Maintains active participation in the learning process