In Part 1, you learned the six fundamentals of prompt thinking — specificity, context-setting, role assignment, constraints, iteration, and output specification. You saw how each one transforms a vague request into a precise communication that produces useful results. Part 2 puts those fundamentals to work. What follows is a library of twelve prompt templates — three for each of the four modes of AI engagement — designed to be copied, adapted, and used immediately.
Your Prompt Library
How to Use These Templates
Each template is a well-constructed prompt that applies the fundamentals from Part 1. The bracketed placeholders — like [your topic] or [describe your challenge] — mark the elements you replace with your own specifics. The structural framework around those placeholders is the part you keep, because it is built from the fundamentals that produce high-quality output.
Think of these templates the way a carpenter thinks of blueprints. The blueprint provides the proven structure. You provide the dimensions, the materials, and the purpose. Together, they produce something that works — not because you followed instructions blindly, but because you combined a reliable structure with your specific needs.
Three Levels of Depth
Within each mode, the three prompts are arranged by depth — simple, intermediate, and advanced. This progression is not about difficulty in the sense of being harder to use. It is about the density of fundamentals in play and the sophistication of the results they produce.
The simple prompts use two or three fundamentals — typically specificity and context-setting — and are designed for straightforward tasks where a clear, focused result is all you need. Start here if you are new to structured prompting. These prompts alone will produce significantly better results than the unstructured queries most people use.
The intermediate prompts add role assignment and constraints, producing more targeted, more nuanced output. They are for situations where you need AI to approach the task from a specific angle or within specific boundaries.
The advanced prompts combine most or all six fundamentals — specificity, context, role, constraints, iteration guidance, and output specification — into a single, layered communication. They are for situations where you need sophisticated, precisely structured output and want to maximize the value of a single AI interaction.
Begin wherever you feel comfortable. Return to the advanced templates as your prompt thinking develops. Over time, the fundamentals will become instinctive, and you will construct prompts at the advanced level without consciously choosing each element — the same way a skilled writer constructs clear sentences without thinking about grammar rules.
You have been assigned a topic for a school essay and need to understand the landscape before you begin writing. You want a clear overview that maps out the key areas so you know where to focus your deeper research.
I need to write an essay on [your topic]. I'm a [your level, e.g., high school junior] and this is my first time studying this subject. Give me a structured overview that covers: the main sub-topics within this subject, the key concepts I need to understand, and the major debates or open questions that would make for a strong essay focus. Keep the overview under 600 words.
This prompt combines specificity (a structured overview with three defined elements), context-setting (your level and familiarity with the subject), and a constraint (under 600 words, preventing an overwhelming response). The result is a focused map of the topic that tells you where to look rather than trying to tell you everything — which is exactly what you need before you begin deeper research. The three elements (sub-topics, key concepts, debates) ensure the overview is organized for action, not just information.
You are exploring a subject that interests you for personal growth or career exploration. You want to understand it deeply enough to decide whether to pursue it further, and you want AI to teach it to you at a level that builds genuine comprehension rather than surface familiarity.
Act as a patient, knowledgeable tutor. I want to understand [your subject]. My current knowledge level: [describe what you already know, even if it's very little]. My goal: [why you want to learn this — career interest, personal curiosity, academic requirement].
Explain the core concepts of this subject in a way I can follow, starting from the fundamentals and building up. Use analogies from everyday life where possible. After each major concept, give me one question I should be able to answer if I truly understood it — do not give me the answer, let me test myself. If I cannot answer, I will ask you to explain further.
This prompt adds role assignment (patient tutor) and builds in an active learning mechanism (self-test questions after each concept). The context is layered — it includes your knowledge level and your goal, allowing AI to calibrate both the starting point and the direction. The instruction to use analogies leverages AI's pattern-matching strength. The self-testing instruction prevents passive consumption — you are forced to engage actively with each concept before moving on. This prompt turns AI from an information dispenser into a structured learning partner.
You are investigating a complex, contested topic where different perspectives offer fundamentally different interpretations. You want to understand the full landscape of views — not just the mainstream position — and you want the output organized for critical analysis rather than passive reading.
Act as a balanced academic researcher presenting a multi-perspective analysis. My topic: [your contested or complex topic]. I am a [your level] student who already understands [what you already know about this topic].
Present the three to four most significant perspectives on this topic. For each perspective: (1) state the core argument in its strongest form, (2) identify the evidence its proponents cite, (3) name one specific weakness that critics of this perspective identify, and (4) note which academic disciplines or cultural traditions tend to hold this view.
Do not present any perspective as correct. Do not default to the most mainstream Western academic view as the baseline — treat all perspectives with equal analytical rigor. After presenting all perspectives, identify the key points of genuine disagreement — not misunderstandings, but substantive differences in values, evidence interpretation, or underlying assumptions.
Format this as a structured comparison, not flowing prose. Use clear headers for each perspective.
This prompt deploys all six fundamentals simultaneously. Role assignment (balanced academic researcher) shapes the analytical stance. Context-setting (your level and existing knowledge) calibrates the depth. Specificity (four-element structure for each perspective) ensures the output is organized for comparison. Constraints (no default to mainstream, no declaring a winner) prevent the sycophancy and bias tendencies identified in Module 1. The instruction to identify genuine disagreements rather than misunderstandings pushes AI toward analytical depth. Output specification (structured comparison with headers, not prose) ensures the response is formatted for critical analysis. This prompt produces a sophisticated, multi-perspective briefing that would take hours of independent research to assemble.
You have a project or assignment and feel stuck on how to approach it. You need a range of ideas to get your thinking started — not a finished plan, but a set of directions to consider.
I'm working on [describe your project or assignment in 1-2 sentences]. I'm stuck on how to approach it. Give me eight different angles I could take, including at least two that are unconventional or unexpected. For each angle, give a one-sentence description and one reason it could be effective.
This prompt uses specificity (eight angles, with format defined) and a constraint (at least two unconventional) that pushes AI past its tendency to generate only the most mainstream ideas. The one-sentence-plus-one-reason format ensures each suggestion is concrete enough to evaluate without being so detailed that it overwhelms. The result is a curated set of starting directions — raw material for your judgment, not a finished answer. Your job after receiving this is to evaluate, select, and develop the most promising direction.
You are facing a real challenge — a decision, a problem, or a situation with multiple dimensions — and you want AI to help you think through it systematically rather than just generate ideas.
Act as a strategic thinking partner. Here is my challenge: [describe your challenge in detail — what you're trying to achieve, what's making it difficult, and any constraints you're working within].
Help me think through this systematically. First, identify the key dimensions of this challenge — what are the different factors I need to consider? Then, for each dimension, suggest two possible approaches with a brief trade-off analysis for each. Do not tell me which approach to choose — present the options clearly so I can make the decision based on my own priorities.
Role assignment (strategic thinking partner) sets AI up for analytical depth rather than quick suggestions. The two-step structure (identify dimensions, then generate options per dimension) produces systematic thinking rather than a flat list. The constraint against recommending a choice prevents sycophancy — AI presents options for your judgment rather than making the decision for you. The trade-off analysis for each option gives you the information you need to evaluate, without collapsing the evaluation into AI's preference. This prompt produces something that resembles a strategic brief — organized, multi-dimensional, and decision-ready.
You have developed an idea, a plan, or a proposal and want rigorous critical feedback — not encouragement, not validation, but a genuine stress-test that identifies weaknesses before you commit resources or submit the work.
Act as a skeptical but fair evaluator — someone who wants my idea to succeed but will not overlook its weaknesses. Here is my [plan / proposal / idea / argument]:
[Present your idea in full detail]
Evaluate this rigorously. Structure your evaluation as follows: (1) Identify the two strongest elements — what is genuinely well-conceived and why. (2) Identify the three most significant weaknesses or risks — be specific about what could go wrong and why. (3) For each weakness, suggest one concrete way it could be addressed or mitigated. (4) Identify one assumption I am making that I may not have examined — something I seem to be taking for granted that could be wrong.
Be direct. Do not soften criticism with excessive praise. I need honest assessment, not encouragement.
This prompt is designed to overcome AI's sycophancy tendency — its default to agree and encourage. The role (skeptical but fair evaluator) and the explicit constraint (do not soften criticism, be direct) push AI toward honest assessment. The four-part structure ensures balanced evaluation: strengths are acknowledged (preventing dismissal), weaknesses are identified with specifics (preventing vagueness), mitigations are suggested (making criticism actionable), and hidden assumptions are surfaced (connecting directly to Module 1 of Layer 1's critical thinking and Module 3 Part 1's assumption work). The instruction to identify unexamined assumptions is particularly powerful — it uses AI to do what the Assumption Audit in Layer 1 did, applied to a specific plan or idea.
You have written a draft of something — an essay, an email, a report — and want AI to help you improve the clarity and flow without rewriting your voice or changing your ideas.
Here is a [type of document, e.g., essay draft, email, report] that I wrote. I want to improve it without losing my voice or changing my main ideas. Please identify: (1) any sentences or passages that are unclear or awkwardly phrased, (2) any places where the structure or flow could be tightened, and (3) any grammatical or punctuation errors. Point me to the specific areas — do not rewrite the entire document for me.
[Paste your draft here]
The critical constraint — "do not rewrite the entire document for me" — keeps the cognitive work on your side. AI identifies the problems; you do the fixing. This preserves your voice, your ownership of the writing, and your skill development. The three-category structure (clarity, structure, grammar) ensures comprehensive feedback without AI taking over. You learn from each revision because you are the one making the changes, guided by AI's targeted identification of where the writing needs work.
You have content in one format — notes, raw data, a rough outline — and need it converted into a different, more polished format for a specific audience, without AI adding or interpreting anything beyond what you provided.
Convert the following [raw notes / meeting notes / data / outline] into a [desired format, e.g., structured summary, professional report, presentation outline, comparison table]. The audience is [describe who will read or see this].
Important constraints: preserve all the original information — do not add facts, opinions, or interpretations that are not in the source material. Do not omit anything unless I have marked it as optional. Maintain the meaning of the original even when reorganizing the structure. If anything in the source material is ambiguous, flag it rather than interpreting it.
[Paste your content here]
This prompt leverages AI's strength in format conversion (a reliable, pattern-based task) while constraining the most common risk: AI adding information or interpretations that were not in your original material. The constraint to flag ambiguity rather than interpret it is particularly valuable — it prevents AI from making decisions on your behalf about meaning. Context-setting through audience specification ensures the format is calibrated for who will actually receive it. You get a clean, reformatted document that is your content in a better container, not AI's content wearing your label.
You have completed a substantial piece of writing — an essay, a research paper, a business proposal — and want a comprehensive editorial review that addresses argument quality, evidence strength, structural coherence, and audience appropriateness, all in a single pass.
Act as a senior editor reviewing a [type of document] before publication. The intended audience is [describe the audience]. The purpose of this document is [what it should accomplish — persuade, inform, propose, analyze].
Review the document below across four dimensions:
(1) Argument quality: Is the central argument clear, logical, and well-supported? Identify any logical gaps, unsupported claims, or places where the reasoning weakens. (2) Evidence: Is the evidence sufficient and appropriately used? Flag any claims that need stronger support or citations. (3) Structure: Does the document flow logically from section to section? Identify any places where the organization is confusing or where transitions are weak. (4) Audience calibration: Is the tone, level of detail, and language appropriate for the intended audience? Identify any places where the writing is too technical, too casual, or misaligned with the reader's likely expectations.
For each issue you identify, be specific: quote the relevant passage, explain the problem, and suggest a direction for improvement. Do not rewrite the passages — give me guidance that I can use to revise them myself. Prioritize the most impactful issues first.
[Paste your document here]
This prompt combines all six fundamentals into a comprehensive editorial review framework. Role assignment (senior editor) sets the standard for rigor. Context-setting (audience and purpose) ensures the review is calibrated to what the document needs to achieve. The four-dimension structure (argument, evidence, structure, audience) ensures nothing is overlooked. The constraint to quote relevant passages, explain problems, and suggest directions — without rewriting — keeps your ownership of the document intact while providing actionable, specific feedback. The prioritization instruction prevents a list of minor issues from burying the major ones. This prompt produces the kind of editorial feedback that professional writers pay for.
You have a long document — an article, an email thread, a report — and need to quickly understand the key points without reading the entire thing.
Summarize the following [article / email / report / document] in exactly three sentences. The first sentence should capture the main point. The second should identify the most important supporting detail or evidence. The third should state any action required or conclusion drawn.
[Paste the text here]
Even a simple utility task benefits from specificity and output specification. The three-sentence constraint with defined content for each sentence (main point, key evidence, action/conclusion) produces a structured summary that is immediately actionable. Compare this to "summarize this article," which produces a vague paragraph of variable length. The structured version takes the same amount of time for AI to generate but is dramatically more useful — because you defined what "useful" means for your purpose.
You need to translate a message or text into another language, but you want the translation to preserve the tone, intent, and cultural nuance — not just convert words mechanically.
Translate the following text into [target language]. The text is a [type of communication, e.g., professional email, casual message to a friend, formal letter, social media post] and the tone is [describe the tone, e.g., warm but professional, casual and friendly, formal and respectful].
Preserve the tone and intent of the original — do not make it more formal or more casual than the source. If any phrases or idioms do not translate directly, adapt them to a culturally equivalent expression in [target language] rather than translating literally. After the translation, note any phrases where you made a cultural adaptation and explain why.
[Paste the text here]
This prompt transforms a mechanical translation task into a nuanced communication task. Context-setting (type of communication and tone) ensures the translation matches the intent, not just the words. The constraint against changing formality level prevents AI from defaulting to a more formal register, which is a common tendency. The instruction to adapt idioms culturally rather than literally produces translations that sound natural in the target language. The request to note cultural adaptations gives you transparency into AI's choices — you can evaluate whether the adaptations are appropriate rather than accepting them blindly.
You have a piece of writing — a cover letter, a personal statement, a professional bio — that needs to be adapted for a specific audience, context, or purpose while preserving your core message and authentic voice.
I need to adapt the following [document type] for a specific context. The original was written for [original audience or purpose]. I now need a version for [new audience or purpose].
Adapt the tone and emphasis to suit the new audience, but do not change the core facts or fabricate new content. Specifically: adjust the level of formality to match [desired formality, e.g., professional but approachable, highly formal, conversational]. Emphasize the aspects most relevant to [what the new audience cares about]. De-emphasize or remove details that are not relevant to this audience.
After the adapted version, provide a brief note explaining the three most significant changes you made and why each one serves the new audience better than the original phrasing.
[Paste your original text here]
This prompt turns a simple adaptation task into a strategic communication exercise. Context-setting is doubled — you specify both the original context and the new context, giving AI the full picture of the transformation required. Constraints (no fabrication, no changing core facts) prevent AI from inventing credentials or experiences. The instruction to explain the three most significant changes gives you editorial transparency — you can see AI's reasoning, evaluate whether the changes are appropriate, and learn from the adaptation logic for future use. The result is not just an adapted document but an education in audience-calibrated communication.
Twelve Templates, Infinite Applications
You now hold something practical and immediately useful: a library of twelve prompt templates that cover the four modes of AI engagement across three levels of depth. Each template is built from the six fundamentals you learned in Part 1 — specificity, context-setting, role assignment, constraints, iteration, and output specification — combined in ways that are calibrated to the task at hand.
But these twelve templates are not the point. They are demonstrations. The real value is in the thinking behind them — the habit of asking, before you type anything: what am I trying to accomplish, what does AI need to know about my situation, what perspective would be most useful, what do I not want, and what format serves my purpose?
As you use these templates — copying them, adapting them, modifying the placeholders for your own needs — you will begin to internalize the patterns. You will start constructing prompts that use the fundamentals naturally, without consciously thinking about which principle you are applying. That is the goal. Not to memorize templates but to develop the instinct for clear, structured communication with AI that produces consistently useful results.
Copy. Adapt. Use. And when a template does not quite fit your situation, build your own — using the same fundamentals, applied to your specific need. The templates are training wheels. The fundamentals are the skill. The skill is yours to keep.