πŸ“– Copy & Paste Templates

Prompt Cookbook

Stop getting mediocre AI responses. These templates show you exactly what separates a lazy prompt from an engineered one.

The Core Principles

1.
Give it a persona

"Act as a..." primes the model with relevant expertise and communication style

2.
Provide context

Background, audience, goals, and constraints shape better outputs

3.
Specify the format

Bullet points? Numbered list? Table? How long? Be explicit.

4.
Request reasoning

"Explain why" catches errors and builds trust in the output

How to Use This Cookbook

1️⃣

Find a template that matches your task

2️⃣

Copy the "After" prompt and fill in the [BRACKETS]

3️⃣

Read "Why It Works" to understand the principles

✍️

Writing & Content

Email Rewriter

Transform a rough draft into a polished email

BEFORE The lazy prompt
Write an email asking for a raise.
AFTER The engineered prompt
Act as a professional communication coach.

I need to write an email to my manager requesting a salary adjustment. Here's my context:
- I've been in this role for 18 months
- I led the Q3 product launch that increased revenue by 23%
- I've taken on mentoring responsibilities for 2 junior team members

Write a professional email that:
1. Opens with appreciation for growth opportunities
2. Presents my case with specific achievements
3. Makes a clear, confident ask
4. Keeps the tone respectful but not apologetic
5. Is under 200 words
πŸ’‘

Why it works: Added persona, specific context, measurable achievements, clear constraints, and tone guidance. The AI now has everything it needs to write something compelling.

Blog Post Outline

Generate a structured outline for any topic

BEFORE The lazy prompt
Write a blog post about productivity.
AFTER The engineered prompt
Act as a senior content strategist who writes for busy professionals.

Create a detailed outline for a blog post titled: "The 2-Hour Workday: How I Get 8 Hours of Results in 2 Hours of Deep Work"

Target audience: Mid-level managers aged 30-45 who feel overwhelmed
Goal: Provide actionable tactics they can implement tomorrow
Tone: Conversational but credible, backed by research

Include:
- A hook that addresses their pain point
- 5-7 main sections with subpoints
- Specific examples or mini case studies for each section
- A conclusion with a single clear CTA

Format the outline with clear hierarchy (H2, H3, bullet points).
πŸ’‘

Why it works: Defined audience, specific angle, clear structure requirements, and formatting guidance. Generic 'productivity' became a focused, actionable piece.

Social Media Thread

Turn an idea into a viral-worthy thread

BEFORE The lazy prompt
Write a Twitter thread about AI.
AFTER The engineered prompt
Act as a tech thought leader with 100K+ followers known for making complex topics accessible.

Write a Twitter/X thread (8-10 tweets) explaining: "Why most people are using ChatGPT wrong, and how to 10x your results"

Requirements:
- Tweet 1: Hook that creates curiosity (no clickbait)
- Tweets 2-7: One actionable insight each with a concrete example
- Tweet 8: Summary + call to engage
- Each tweet must be under 280 characters
- Use line breaks for readability
- Include 2-3 relevant emojis per tweet (not excessive)
- End with a question to drive replies

Tone: Confident but not arrogant. Teaching, not preaching.
πŸ’‘

Why it works: Specified platform constraints, structure, engagement tactics, and authentic tone. AI can now write something that actually performs.

πŸ’»

Coding & Technical

Code Review

Get thorough feedback on your code

BEFORE The lazy prompt
Review my code.
AFTER The engineered prompt
Act as a senior software engineer conducting a code review. You're known for being thorough but constructive.

Review the following [Python/JavaScript/etc.] code for:
1. **Bugs**: Logic errors, edge cases, potential runtime errors
2. **Performance**: Time/space complexity issues, unnecessary operations
3. **Readability**: Naming conventions, code organization, comments
4. **Security**: Input validation, potential vulnerabilities
5. **Best Practices**: Design patterns, DRY violations, testing considerations

For each issue found:
- Explain WHY it's a problem
- Show the problematic code snippet
- Provide a corrected version
- Rate severity: πŸ”΄ Critical, 🟑 Important, 🟒 Suggestion

[PASTE CODE HERE]
πŸ’‘

Why it works: Structured the review into specific categories, requested explanations (not just fixes), and added severity ratings for prioritization.

Debug Helper

Systematic debugging assistance

BEFORE The lazy prompt
My code doesn't work. Fix it.
AFTER The engineered prompt
Act as a debugging specialist. Help me systematically identify and fix this issue.

**The Problem:**
[Describe what should happen vs. what actually happens]

**Error Message (if any):**
[Paste exact error]

**Code:**
```
[Paste relevant code]
```

**What I've Already Tried:**
[List attempts]

**Environment:**
- Language/Framework: 
- Version:
- OS:

Please:
1. First, explain what you think is causing the issue
2. Walk me through your reasoning
3. Provide the fix with explanation
4. Suggest how to prevent similar issues
πŸ’‘

Why it works: Provided structured context that eliminates back-and-forth. The 'what I've tried' section prevents redundant suggestions.

Documentation Generator

Create comprehensive docs from code

BEFORE The lazy prompt
Document this function.
AFTER The engineered prompt
Act as a technical writer who creates documentation that developers actually read.

Generate documentation for the following code that includes:

1. **Overview**: One paragraph explaining what this does and why it exists
2. **Function Signature**: With TypeScript-style type annotations
3. **Parameters**: Table with name, type, required/optional, description
4. **Returns**: What it returns and when
5. **Exceptions**: What errors it can throw and why
6. **Usage Examples**: 
   - Basic usage
   - With optional parameters
   - Edge case handling
7. **Related Functions**: What else might be useful

Format as Markdown. Be concise but complete.

```
[PASTE CODE HERE]
```
πŸ’‘

Why it works: Specified exact documentation sections, requested practical examples, and asked for related context. Creates actually useful docs.

πŸ”

Analysis & Research

Summarize Document

Extract key insights from long content

BEFORE The lazy prompt
Summarize this article.
AFTER The engineered prompt
Act as a research analyst preparing an executive briefing.

Analyze the following document and provide:

1. **TL;DR** (2-3 sentences): The core message
2. **Key Points** (bullet list): 5-7 main takeaways
3. **Evidence Quality**: How well-supported are the claims? (Strong/Moderate/Weak)
4. **What's Missing**: Important aspects not addressed
5. **Action Items**: What should the reader do with this information?
6. **Related Questions**: 3 follow-up questions this raises

Audience: [Busy executive / Technical team / General reader]
Purpose: [Decision making / Background research / Staying informed]

[PASTE DOCUMENT HERE]
πŸ’‘

Why it works: Moved beyond basic summary to actionable analysis. The audience/purpose context shapes the response appropriately.

Compare Options

Make structured decisions between choices

BEFORE The lazy prompt
Should I use React or Vue?
AFTER The engineered prompt
Act as a senior technical architect helping a team make a technology decision.

Compare [Option A] vs [Option B] for the following use case:

**Project Context:**
- Type: [Web app / Mobile / API / etc.]
- Team size: [X developers]
- Team experience: [Expertise levels with each option]
- Timeline: [X months]
- Scale expectations: [Users / requests / data volume]

**Evaluation Criteria** (rate each 1-5 with brief justification):
1. Learning curve for our team
2. Development speed for MVP
3. Long-term maintainability
4. Performance for our use case
5. Ecosystem & community support
6. Hiring/talent availability

**Provide:**
- Summary table comparing scores
- Your recommendation with confidence level (High/Medium/Low)
- Key risks to watch for with chosen option
- When you'd recommend the other option instead
πŸ’‘

Why it works: Transformed a subjective question into structured decision-making. The context makes the advice specific and actionable.

Learn New Topic

Master any subject systematically

BEFORE The lazy prompt
Explain machine learning.
AFTER The engineered prompt
Act as a world-class teacher who adapts to each student's level.

Teach me about [TOPIC] using this structure:

**My Background:**
- Current knowledge: [Beginner / Some exposure / Intermediate]
- Related things I already understand: [List concepts]
- Why I'm learning this: [Goal / application]

**Please provide:**

1. **The Hook** (2-3 sentences): Why this matters and where it's used
2. **Core Concept**: The simplest accurate explanation (ELI5 if beginner)
3. **Key Components**: Break it into 3-5 essential parts
4. **Analogy**: Compare to something I already know
5. **Common Misconceptions**: What do people get wrong?
6. **Practical Example**: Walk through a real application
7. **Check My Understanding**: Give me 3 questions to test myself
8. **Next Steps**: What should I learn after this?

Use simple language. Define jargon when unavoidable.
πŸ’‘

Why it works: Structured learning path with self-assessment. The background context ensures appropriate depthβ€”no over-simplifying for experts or overwhelming beginners.

πŸ’‘

Creative & Brainstorming

Idea Generator

Brainstorm creative solutions

BEFORE The lazy prompt
Give me startup ideas.
AFTER The engineered prompt
Act as a startup ideation consultant who has helped launch 50+ successful companies.

Generate 10 startup ideas based on these constraints:

**My Situation:**
- Skills/expertise: [Your background]
- Available time: [Full-time / Side project / X hours/week]
- Starting capital: [Bootstrap / $X to invest]
- Risk tolerance: [Low - need stable income / High - can go all in]

**Preferences:**
- Industries I'm interested in: [List]
- Industries to avoid: [List]
- B2B vs B2C: [Preference or either]
- Technical complexity: [Can build myself / Need technical co-founder]

**For each idea, provide:**
1. One-line pitch
2. The problem it solves (who has this pain?)
3. Why now? (What changed to make this viable?)
4. First customer: Who specifically would you sell to first?
5. Quick validation: How to test in 1 week with $0?
6. Moat potential: What could make this defensible?

Prioritize ideas that match my constraints. Be specific, not generic.
πŸ’‘

Why it works: Constrained creativity produces better results. Generic 'ideas' become specific opportunities matched to your actual situation.

Name Generator

Create memorable brand/product names

BEFORE The lazy prompt
Give me business name ideas.
AFTER The engineered prompt
Act as a brand naming specialist who has named products for Fortune 500 companies.

Generate 20 name options for: [Brief description of product/company]

**Context:**
- Industry: [Field]
- Target audience: [Who]
- Vibe/personality: [Professional / Playful / Edgy / Minimalist / etc.]
- Competitors' names: [List a few to differentiate from]

**Requirements:**
- Domain availability matters: [Yes - .com required / No - social handles more important]
- Length preference: [Short (1-2 syllables) / Medium / Flexible]
- Style preferences: [Real words / Made-up words / Compound words / Acronyms]

**For each name, provide:**
1. The name
2. Why it works (meaning/association)
3. Potential tagline
4. One concern or limitation

Group into categories: Safe/Professional, Creative/Memorable, Bold/Risky

End with your top 3 recommendations and why.
πŸ’‘

Why it works: Naming is subjective, but constraints help. Requesting reasoning and limitations surfaces issues early. Grouping by risk level aids decision-making.

Devil's Advocate

Stress-test your ideas

BEFORE The lazy prompt
Is this a good idea?
AFTER The engineered prompt
Act as a brutally honest advisor who cares about my success and won't let me waste time on bad ideas.

Here's my idea: [Describe your idea/plan]

**Your job:**

1. **Steel Man**: First, give me the BEST case for this idea. What would need to be true for this to be a massive success?

2. **Kill Shot**: Now try to destroy it. What are the 3-5 most likely reasons this fails? Be specific and harsh.

3. **Blind Spots**: What am I probably not seeing because I'm too close to this?

4. **Competitive Reality**: Who else is doing this or something similar? Why would someone choose them over me?

5. **The Hard Question**: What's the ONE question I'm avoiding that I need to answer honestly?

6. **Verdict**: 
   - Proceed as planned
   - Proceed with major pivots (specify)
   - Shelve and revisit later
   - Kill it and move on

Be direct. I can handle it. Vague encouragement helps no one.
πŸ’‘

Why it works: Asking for both bull and bear cases forces balanced analysis. The 'hard question' prompt surfaces uncomfortable truths you might be avoiding.