The Prompt Bible

The complete guide to getting exactly what you want from AI.

About This Book

This book teaches you how to write prompts for an AI chat assistant so that the answer you get is the answer you needed. It covers how these assistants really work, how to build a strong prompt, how to shape and repair answers, twelve reusable patterns, how to reduce made-up content, how to work with your own material safely, prompts for eight kinds of everyday work, how to build your own prompt systems, and how to use AI responsibly. More than 150 prompts are written out in full, ready to copy.

These are starting prompts for use with a general-purpose AI chat tool of your choice (ChatGPT, Claude, Gemini, or similar) - a separate account/subscription with that provider is required and not included. Output quality depends on the model you use and the details you fill in; review any AI output before relying on it, especially for legal, financial, or medical claims.

How to Read It

• Fast start (about an hour): read Chapters 2, 3, 4 and 8, then keep Appendix B beside you.
• Full course: read in order and follow the 14-day practice plan in Appendix E.
• By task: go straight to the chapter in Part 6 that matches your work, then return to Parts 1 and 2 when something does not behave.

Conventions

• Every prompt appears in a code block so you can copy it. Words in square brackets, such as [READER], are placeholders to replace with your own details.
• Names, numbers and scenarios in examples are illustrative. Use your own.
• AI tools change quickly. Where a feature, limit or specification matters, the book gives a typical value and tells you to check the app. Statements about tools reflect the situation as of late 2026.
• AI output is a draft. You are responsible for checking it before you use it.

Contents

• Part 1: Foundations
  • Chapter 1: What an AI Chat Model Really Is, and Why It Behaves the Way It Does
  • Chapter 2: The Seven Parts of a Strong Prompt
  • Chapter 3: The Specificity Ladder
• Part 2: Getting Exactly What You Want
  • Chapter 4: Define "Done" First
  • Chapter 5: Format Control
  • Chapter 6: Show, Don't Tell
  • Chapter 7: Iteration
  • Chapter 8: When the Answer Is Wrong
• Part 3: Prompt Patterns
  • Chapter 9: Twelve Reusable Patterns
• Part 4: Thinking and Quality
  • Chapter 10: Making It Reason
  • Chapter 11: Reducing Made-Up Answers
  • Chapter 12: Long Documents and Big Tasks
• Part 5: Working With Your Own Material
  • Chapter 13: Summaries, Extraction and Comparison
  • Chapter 14: Spreadsheets and Data in Plain Language
  • Chapter 15: Privacy
• Part 6: Prompting by Task
  • Chapter 16: Writing and Editing
  • Chapter 17: Business and Marketing
  • Chapter 18: Learning and Studying
  • Chapter 19: Research and Decision-Making
  • Chapter 20: Coding Help
  • Chapter 21: Planning and Productivity
  • Chapter 22: Creative Work
  • Chapter 23: Image, Audio and Video Prompting Basics
• Part 7: Building Systems
  • Chapter 24: Your Personal Prompt Library
  • Chapter 25: Custom Instructions, System Prompts and Reusable Assistants
  • Chapter 26: Prompt Chains and Simple Workflows
  • Chapter 27: Testing and Improving Prompts
  • Chapter 28: AI Agents and Tools: What They Add and the Risks
• Part 8: Responsible Use
  • Chapter 29: Honesty, Disclosure, Copyright, Bias and Knowing When Not to Use AI
• Appendices
  • Appendix A: 60 Before-and-After Prompt Rewrites
  • Appendix B: One-Page Cheat Sheet
  • Appendix C: Troubleshooting Table
  • Appendix D: Glossary of 60 Terms in Plain English
  • Appendix E: A 14-Day Practice Plan

Part 1: Foundations

Before you learn techniques, you need a working picture of the thing you are talking to. Most bad prompts come from a wrong mental model: people treat an AI chat assistant like a search engine, a mind reader, or a human colleague who remembers last week. It is none of those. Two short chapters fix that, and then you will have a reliable structure to build every prompt on.

Chapter 1: What an AI Chat Model Really Is, and Why It Behaves the Way It Does

After this chapter you will be able to: explain in plain words why an AI assistant produces what it produces, predict where it is likely to go wrong, and adjust how you write prompts to match.

It predicts text from patterns

An AI chat assistant is a program trained on a very large amount of text so that it can predict what text should come next. When you send a prompt, it does not look up an answer in a filing cabinet. It generates a reply one small piece at a time, and each piece is chosen because it fits well with everything that came before it. Those small pieces are usually called tokens. A token is often a short word or part of a word. As a rough rule of thumb, one token is about three quarters of an English word, so a 1,000-word document is typically somewhere around 1,300 tokens. Treat that as an estimate, not a measurement.

This single fact explains most of what you will see:

• It is strong at anything that resembles patterns it has met many times: emails, summaries, outlines, explanations, rewrites in a different tone, code in popular languages, lists of ideas.
• Fluent text is cheap for it. Truth is not produced automatically. It generates what is plausible, which often overlaps with what is true but is not the same thing.
• Your prompt is the opening of the text it continues. A clearer, richer opening usually leads to a better continuation. Everything in this book is a method for writing a better opening.

The context window: its entire working desk

...