ChatGPT & AI Prompt Engineering Cheat Sheet

Last Updated: November 21, 2025

Core Prompting Techniques

Technique Description Example
Zero-Shot Direct instruction without examples Translate this to French: [text]
Few-Shot Provide 2-5 examples first Example 1: input → output
Example 2: input → output
Now: [your input]
Chain-of-Thought Ask to show reasoning steps Let's think step by step...
Role Prompting Assign a specific persona You are an expert Python developer...
Instruction Following Clear numbered steps 1. Analyze 2. Summarize 3. Recommend
Constraint-Based Set specific limitations In exactly 100 words, explain...
Tree of Thoughts Explore multiple reasoning paths Consider 3 different approaches...
Self-Consistency Generate multiple answers, choose best Solve this 3 times and pick most consistent

Writing Prompt Templates

Task Template
Blog Post Write a [word count] blog post about [topic] for [audience]. Include: 1) Hook opening, 2) 3 main points with examples, 3) Call-to-action. Tone: [casual/professional/technical]
Email Draft Draft a [formal/casual] email to [recipient] about [topic]. Purpose: [inform/request/apologize]. Keep under [word count] words.
Social Media Create [number] [platform] posts about [topic]. Include: relevant hashtags, engaging hook, call-to-action. Character limit: [limit]
Product Description Write a compelling product description for [product]. Highlight: [benefits]. Target audience: [demographic]. Include emotional appeal and specifications.
Resume Bullet Transform this experience into a strong resume bullet: [description]. Use action verbs, quantify results, highlight impact.
Story Writing Write a [genre] story about [premise]. Setting: [location/time]. Include: character development, conflict, resolution. Length: [words]
Academic Paper Write a [word count] academic paragraph on [topic]. Include: thesis statement, 3 supporting arguments with citations, counter-argument, conclusion.

Coding Prompt Templates

Task Template
Code Generation Write a [language] function that [task]. Requirements: [specs]. Include error handling and comments. Optimize for [speed/readability].
Code Review Review this code for: 1) Bugs, 2) Performance issues, 3) Security vulnerabilities, 4) Best practices. Suggest improvements: [code]
Debugging Debug this [language] code. Error: [error message]. Expected behavior: [description]. Code: [code]
Refactoring Refactor this code to be more [maintainable/efficient/readable]. Apply [design patterns]. Preserve functionality: [code]
Documentation Generate comprehensive documentation for this [language] code. Include: function descriptions, parameters, return values, examples: [code]
Unit Tests Write unit tests for this function using [framework]. Cover: edge cases, error conditions, normal cases: [code]
Algorithm Explanation Explain this algorithm step-by-step at a [beginner/intermediate/advanced] level. Include time/space complexity: [code]

Analysis & Research Templates

Task Template
Data Analysis Analyze this data: [data]. Find: patterns, outliers, correlations. Provide: insights, visualizations (describe), recommendations.
SWOT Analysis Conduct a SWOT analysis for [topic/company]. Consider: market position, competitors, trends, resources.
Pros/Cons Compare [option A] vs [option B] for [use case]. Analyze: cost, ease of use, scalability, maintenance. Recommend best option.
Market Research Research [industry/product]. Cover: market size, key players, trends, opportunities, threats, target audience.
Summarization Summarize this in [length]: [text]. Focus on: key points, actionable insights, main conclusions. Format: [bullets/paragraph]
Competitive Analysis Compare [competitors] in [industry]. Evaluate: features, pricing, market position, strengths, weaknesses.

Creative & Brainstorming Templates

Task Template
Idea Generation Generate 10 creative ideas for [topic]. Criteria: [innovative/practical/low-cost]. Target: [audience]. Include brief description for each.
Problem Solving I have this problem: [problem]. Suggest 5 solutions considering: constraints [list], resources [list], timeline [time].
Name Generation Generate [number] names for [product/company/project]. Style: [modern/classic/playful]. Must: be memorable, relevant to [industry], available as domain.
Campaign Ideas Create [number] marketing campaign ideas for [product]. Target: [audience]. Include: theme, channels, key message, call-to-action.
Content Calendar Create a [timeframe] content calendar for [platform/blog]. Topics: [themes]. Include: titles, format, posting schedule.

Advanced Prompt Structures

Structure Format
RTF Framework Role: You are [role]
Task: [specific task]
Format: Deliver output as [format]
COAST Framework Context: [background info]
Objective: [goal]
Actions: [steps to take]
Scenario: [specific situation]
Task: [deliverable]
RISEN Framework Role: [persona]
Input: [data provided]
Steps: [process]
Expectation: [desired outcome]
Narrowing: [constraints]
TAG Framework Task: [what to do]
Action: [how to do it]
Goal: [desired result]
CARE Framework Context: [situation]
Action: [what AI should do]
Result: [expected outcome]
Example: [sample]

Best Practices

Practice Description
Be Specific Use precise language, avoid ambiguity. Include details about format, length, style, audience.
Provide Context Give background information. AI performs better with relevant context about your situation.
Use Delimiters Separate sections with triple quotes """, XML tags, or clear markers for structured input.
Iterate Refine prompts based on output. Ask for revisions: "Make it more technical" or "Simplify this".
Specify Format Request tables, bullet points, code blocks, JSON, or other formats explicitly.
Set Constraints Define word limits, required elements, things to avoid, tone, complexity level.
Request Verification Ask AI to double-check work: "Review your answer for accuracy" or "Show your calculations".
Break Complex Tasks Split large requests into smaller, sequential prompts for better results.

Prompt Modifiers

Modifier Type Examples
Tone professional, casual, friendly, authoritative, empathetic, humorous, formal, conversational
Style concise, detailed, technical, simplified, academic, journalistic, creative, analytical
Audience beginners, experts, children, executives, technical users, general public, [specific age]
Format bullet points, numbered list, table, essay, report, email, code, JSON, markdown
Length 50 words, 500 words, 2 paragraphs, one page, brief summary, comprehensive analysis
Perspective first person, third person, objective, subjective, from [role] viewpoint

Common Pitfalls to Avoid

Pitfall Why It Fails Better Approach
Too vague "Tell me about AI" "Explain how transformer models work in 200 words for beginners"
Missing context "Fix this code" "Debug this Python code that should sort a list but returns empty: [code]"
No format specified "Compare these tools" "Create a comparison table of [tools] with columns: features, pricing, pros, cons"
Overloading Asking 5+ things at once Break into separate prompts or number requests clearly
Assuming knowledge Using niche jargon without explanation Define terms or request clarification if needed
💡 Pro Tips:
  • Start prompts with action verbs: "Create", "Analyze", "Compare", "Generate", "Explain"
  • Use temperature control if available: lower (0.3) for factual, higher (0.8) for creative
  • Ask AI to ask YOU questions: "What additional information do you need to complete this task?"
  • Request multiple options: "Give me 3 different approaches" then choose the best
  • Use the "Act as if..." technique: "Act as if you're a senior architect reviewing this design"
  • Chain prompts: Use output from one prompt as input for the next for complex workflows
  • Save successful prompts as templates for reuse and refinement
  • Test different phrasings of the same request to see which produces better results
  • For code: Always specify the programming language, version, and framework
  • Add "Explain your reasoning" to get insights into AI's decision-making process
← Back to Data Science & ML | Browse all categories | View all cheat sheets