Every lesson, one tool

Prompt lessons for Claude

XML-style tags — <context>, <task>, <format> — are the idiomatic way to separate the parts of a prompt here, and the single biggest practical difference from the other three. Put long reference material first and your instruction last.

  1. 01 The Principle of the Definite Chief Aim A prompt gets you what you actually asked for, so the useful skill is naming the exact thing you want: who it is for, how long it should be, what it must cover, and what you will do with it.
  2. 02 The Blueprint of Creation Instead of describing what you want and hoping the shape comes out right, hand over the empty shape itself - headings, fields, table columns - with square-bracket notes saying what goes in each slot.
  3. 03 The Power of Organized Knowledge (Context) A model knows a great deal about the world and nothing at all about your situation, so generic advice is the only honest thing it can give you until you supply the facts.
  4. 04 The Law of the Master Persona Telling the model who to be changes what it notices.
  5. 05 The Secret of Silent Suggestion (Few-Shot Prompting) Instead of describing the output you want, show two or three finished examples and let the model infer the pattern.
  6. 06 The Erection of Mental Fences (Delimiters) A delimiter is a marker - a row of hashes, a fenced block, an XML tag - that tells the model where your instructions end and your raw material begins.
  7. 07 The Virtue of Persistent Effort (Iteration) A first prompt is a first draft.
  8. 08 The Discipline of Precise Limitation Constraints are the limits you set on the answer before you get it - how long, what format, what tone, and crucially what to leave out.
  9. 09 The Ladder of Logical Progression (Step-by-Step Thinking) Start with the honest part: adding the phrase "think step by step" to a prompt does very little on current models, because they already reason before answering.
  10. 10 Commence with the End in Mind Describe the finished thing before you ask for it - its length, its shape, its tone, and the test it has to pass - rather than describing the activity you want performed.
  11. 11 The Art of the Open Question A question that can be answered yes or no will be answered yes or no.
  12. 12 The Council of Invisible Counselors Instead of asking for an answer, ask several specified viewpoints to argue the question out and then reconcile them.
  13. 13 The Search for Absolute Truth (Fact-Checking) Treat a confident answer as a draft, not a result.
  14. 14 The Transmutation of Form Content and form are separable.
  15. 15 The Power of the Negative Command Naming what you do not want removes a whole class of wrong answers at once - the slang, the tourism paragraph, the mention of flooding.
  16. 16 The Seed of a New Idea (Brainstorming) Brainstorming with an AI only works when you give it somewhere to push against.
  17. 17 The Language of Simplicity A prompt is an instruction, not a conversation.
  18. 18 The Principle of Controlled Imagination Creativity needs a spark and a fence.
  19. 19 The Elimination of Mental Static (Cleanliness) A typo in a technical term, a half-named person, a sentence whose grammar hides two requests inside one - each makes the model guess, and a guess is where a wrong answer comes from.
  20. 20 The Strategic Use of Temperature Temperature is the setting that controls how adventurous a model's word choices are: low makes it predictable and literal, high makes it surprising and inventive.
  21. 21 The Master's Review Before you act on an answer - or before you even send a long prompt - you ask the model to review the work: to find the ambiguities in your instruction, the contradictions you did not notice, and the weak spots in its own draft.
  22. 22 The Chain of Thought Instead of asking for the whole result in one prompt, you break the work into a sequence of linked steps and feed each answer into the next question.
  23. 23 The Reversal of Perspective When an answer misses, instead of rewriting your prompt and trying again, you ask the model what it thought you were asking for.
  24. 24 The Treasure of Specificity A vague request gets an average answer, because the model has to pick a middle of the road you did not describe.
  25. 25 The Formation of the Master Mind The last technique is not a new one - it is the habit of using the others together.