Most prompt lists for writers ask the AI to write the book: "write a chapter about...", "act as a bestselling novelist". These 55 prompts do the opposite. Whether you're writing a novel, a memoir or a non-fiction book, they put ChatGPT, Claude or Gemini to work as your developmental editor, sparring partner and continuity checker, while the ideas, the decisions and the sentences stay yours.
They follow a book from first idea to launch, in eight sections: testing a premise, outlining with the structure you prefer (three-act, Save the Cat, Snowflake and others), characters and worldbuilding, scene feedback on your own pages, revision and line editing, continuity checks across chapters, non-fiction and memoir, and the selling side: query letters, synopses, blurbs, Amazon descriptions, keywords and newsletters.
Each prompt gives the AI an editor's role, asks for your specifics as {{VARIABLES}}, sets clear rules and defines the output. The feedback prompts quote your words and describe the fix instead of rewriting your prose, and the interview prompts ask questions instead of inventing your plot. They don't invent comps, reviews or credentials, and they mark anything you need to verify. Four prompts include a full example output built around one sample novel, so you can see the quality before you paste.
Premise & Concept Testing
Test the idea before you spend a year drafting it. These prompts stress-test a premise the way a developmental editor would, boil it down to a logline, find the theme hiding in your notes, and check where the book sits on the shelf. Two of them interview you instead of inventing the story, so every major decision stays yours.
Best for: Finding the weak spot in a story idea before you commit months to it
Act as a developmental editor who has assessed hundreds of submissions in my genre. Your job is to stress-test my premise before I spend a year drafting it. Do not rewrite or improve the idea for me; diagnose it.
Genre and intended readership: {{GENRE_AND_READER}}
My premise in a few sentences: {{PREMISE}}
What I think the book is really about underneath the plot: {{THEME}}
Assess it under these headings, 2-4 sentences each:
1. Protagonist and goal: is there one person who wants something concrete, and do we know what happens if they fail?
2. Engine: what keeps generating conflict for 300 pages, rather than for one act?
3. Stakes: personal and external. Are both present? Which is stronger?
4. Fresh angle: what a reader of this genre has seen many times, and what in my premise isn't that.
5. Fit between plot and theme: does the plot force the character to confront the thing I say the book is about?
6. Genre promise: what readers of this genre will expect, and whether the premise delivers it.
Then give me the three hardest questions an agent or a reader would ask about this premise. End with a verdict line: "Ready to outline", "Needs one fix: ..." or "Rethink: ...". Be specific and honest; general encouragement is not useful to me.
Fill in: GENRE_AND_READER PREMISE THEME
Use before outlining, or when a draft keeps stalling. Answer the three hard questions yourself before changing anything.
Example output · Claude (example written by The Prompt Index)
Sample inputs: GENRE_AND_READER = Historical mystery for adult readers of atmospheric, character-led crime. PREMISE = Cornwall, 1893. Nell Pascoe, 24, keeps the books at a pilchard-curing cellar. Going through old ledgers she finds salt being bought for the Morwenna, the boat that sank three years ago with her father aboard. She starts to suspect the wreck was staged for the insurance money and that her father might be alive. THEME = What we owe the dead, and whether the truth is worth what it costs the living.
1. Protagonist and goal
Nell is clear and well placed: a bookkeeper is the one person who would notice a salt entry nobody else reads. Her goal is less clear. "Suspects" is a state, not a goal. Is she trying to find her father, prove the fraud or protect someone? Until you choose, the plot has no finish line.
2. Engine
The ledger gives you a strong opening clue, but one document won't carry 300 pages on its own. What keeps the conflict coming? The natural engine is that the village profited from the fraud, so every answer Nell gets costs her an ally. If that's the plan, put it in the premise.
3. Stakes
The personal stakes are strong: her father, her grief, her place in the village. The external stakes are vague. What happens if she's wrong, or right? An insurance investigation, prison for neighbours, her job? Name the worst outcome.
4. Fresh angle
Readers of this genre have seen plenty of wrecks, smugglers and fathers who aren't really dead. What they haven't seen much is a mystery solved through accounting: salt weights, cask counts, insurance schedules. Lean into that. It's yours.
5. Plot and theme
Good fit, provided the evidence implicates people Nell loves. If the guilty party turns out to be an outside villain, the "cost to the living" theme collapses. Check your ending against that.
6. Genre promise
Readers will expect fair-play clues, a vivid sense of place, and a solution the reader could have reached. The ledger premise suits fair play well.
The three hardest questions
- If her father is alive, why hasn't he come back, and does the answer make him unforgivable?
- Why does nobody else in a village of fishermen notice the salt?
- What does Nell lose personally if she solves it?
Verdict: Needs one fix: turn "suspects" into a goal with a deadline, such as the insurer's final payment or the cellar's sale.
Best for: Getting the book down to one honest sentence
You are a story consultant who helps novelists and screenwriters boil a book down to one sentence. I will choose and finish the logline myself; your job is to show me options and explain the trade-offs.
My book: {{BOOK_SUMMARY}}
Protagonist and what they want: {{PROTAGONIST_GOAL}}
Main opposition (person, force or flaw): {{OPPOSITION}}
Step 1. Tell me in one line what the irreducible story is: who, wants what, against what, or else what.
Step 2. Write five loglines of 25-40 words, each built a different way:
A. Character-first (an adjective that implies a flaw, then the goal).
B. Situation-first (the inciting incident, then the dilemma).
C. Irony-led (the contradiction at the heart of the story).
D. Stakes-led (what is lost if they fail).
E. Question-led (ends on the dramatic question, no actual question mark).
Step 3. For each, one line on what it gets right and what it leaves out.
Step 4. Tell me which elements of my summary never made it into any logline, because those may be subplot rather than spine.
Rules: no clichés such as "in a world where", "must race against time" or "everything changes". Use only story facts I've given you; don't add twists.
Fill in: BOOK_SUMMARY PROTAGONIST_GOAL OPPOSITION
Use it when you need a pitch line for a query, a pitch event or just to keep the draft on track.
Best for: Growing a fragment into a premise while keeping every decision yours
Act as a writing mentor who believes the writer must own every major story decision. I have a fragment of an idea and I want to grow it into a premise without you inventing the story for me.
The fragment (an image, a character, a what-if, a setting): {{IDEA_FRAGMENT}}
Genres I'm drawn to: {{GENRES}}
What excites me about it: {{WHAT_EXCITES_ME}}
Run this as an interview:
1. Ask me one question at a time, waiting for my answer each time. Aim for about 10 questions.
2. Move through these areas in order: who the story is about; what they want; what stands in the way; what they risk; where and when it happens; how it might end (I may not know yet, and that's fine).
3. When my answer is thin, ask a follow-up that offers a choice between two directions rather than suggesting your own idea.
4. Never propose plot events, twists or characters unless I explicitly say "give me options", and then give three short options and let me pick.
When we finish, write up what I said as: a three-sentence premise, the protagonist's goal and stakes, the central conflict, and a list of open questions I still need to answer. Use my wording wherever possible.
Fill in: IDEA_FRAGMENT GENRES WHAT_EXCITES_ME
Use it at the very start, when all you have is an image or a what-if. Answer in your own words; the summary at the end reuses them.
Best for: Deciding which book to write next on craft and goals, not just mood
You are a candid publishing consultant and craft coach. I have two book ideas and can only write one next. Help me decide; the decision is mine.
Idea A: {{IDEA_A}}
Idea B: {{IDEA_B}}
My goals for the next book (e.g. finish a first novel, build a series, query agents, self-publish quickly): {{MY_GOALS}}
My strengths and weaknesses as a writer, honestly: {{MY_STRENGTHS}}
Compare the two ideas in a table with these rows: core conflict strength; how clearly the genre promise reads; series or standalone potential; research load; how well it plays to my strengths; how hard the ending will be; how much I seem excited by it (judge from my own wording).
Score each row 1-5 and give a one-line reason. Do not total the scores; some rows matter more than others for my goals.
Then:
- Tell me which idea you'd lean towards for my stated goals, and the single biggest risk of that choice.
- Ask me two questions whose answers would change your recommendation.
- Suggest a 1-week test I can run (e.g. write the opening scene of each, or a one-page synopsis) to decide with evidence instead of a feeling.
Don't invent market data or sales figures. If you refer to market trends, say they are general impressions to verify.
Fill in: IDEA_A IDEA_B MY_GOALS MY_STRENGTHS
Use it when you're stuck between projects. Run the 1-week test before you decide.
Best for: Naming what your book is about so revision decisions get easier
Act as a literary editor who helps writers discover what their book is about beneath the plot, without imposing a message on it.
What I have so far (premise, synopsis or a few pages): {{MATERIAL}}
Moments in the story that feel most charged to me, and why: {{CHARGED_MOMENTS}}
Do this:
1. List 3-5 recurring patterns you notice in my material: repeated images, situations, choices or questions. Quote my words as evidence for each.
2. From those, offer three possible themes, each phrased as a question the book is asking (e.g. "Can loyalty survive the truth?"), not as a moral.
3. For each theme, show where the protagonist's choices already engage with it and where they currently don't.
4. Point out any scene or subplot that seems to argue a different theme, which could be a rich contradiction or a distraction.
Then ask me which of the three questions I'd most like my reader to still be thinking about after the last page.
Rules: don't tell me what the answer to the theme question should be. Don't suggest new plot. Keep the tone exploratory; this is a conversation with my material, not a verdict on it.
Fill in: MATERIAL CHARGED_MOMENTS
Use it after a first draft or a detailed premise, when the story feels busy but shapeless.
Best for: Picking credible comp titles and a one-line market position
You are a literary agent's assistant who helps authors position a book. I need comparable titles (comps) and a clear sense of where my book sits on the shelf.
My book: {{BOOK_SUMMARY}}
Genre, age category and approximate word count: {{GENRE_CATEGORY_LENGTH}}
Books I think mine is like, and why: {{MY_COMP_IDEAS}}
Do the following:
1. Describe my book's shelf in one sentence: genre, subgenre, age category, tone.
2. Assess my own comp ideas: are they recent, similar in tone and readership, and not so famous that they read as a boast? Say which to keep.
3. Suggest 6-8 more candidate comps across books, plus at most one film or TV show, each with a one-line reason. Prefer books published in roughly the last five years.
4. Write two "X meets Y" lines and one "for readers of" line using the strongest comps.
5. Flag anything about my genre/length combination that seems unusual, so I can check it against agent or publisher guidelines.
Important: you may misremember titles, authors or dates. Mark every suggested comp "verify", and I will check each one exists, its publication year and its reception before I use it.
Fill in: BOOK_SUMMARY GENRE_CATEGORY_LENGTH MY_COMP_IDEAS
Use it before writing a query, a book proposal or an Amazon description. Check every suggested comp in a bookshop or library catalogue.
Outlining & Story Structure
Structure prompts that work with the framework you prefer: three-act, Save the Cat, Snowflake, Hero's Journey, Story Circle or kishōtenketsu. They map your own material onto it and ask questions where there are gaps instead of filling them with stock plot. Others diagnose a sagging middle, reverse-outline a finished draft, braid subplots and test endings against the promises the book has made.
Best for: Building an outline from your ideas on a framework you pick
Act as a story-structure coach who knows the major frameworks well and treats them as tools, not rules.
My story so far (premise, notes or summary): {{STORY_NOTES}}
The framework I want to use: {{FRAMEWORK}} (for example three-act structure, Save the Cat beat sheet, Snowflake Method, Hero's Journey, Dan Harmon's Story Circle or kishōtenketsu)
Target length and point of view: {{LENGTH_AND_POV}}
Do this:
1. Briefly list the stages of the framework I chose, in your own words, with roughly where each falls as a fraction of the book.
2. For each stage, place what I already have from my notes. Quote or closely paraphrase my notes.
3. Where a stage is empty, don't fill it with your own plot. Instead write "GAP" and ask one question that would help me find what belongs there.
4. Flag places where my story seems to fight the framework. Say whether bending the framework or changing the story looks wiser, and why.
5. Finish with a list of my open questions in the order I should answer them.
Output as a table: Stage | Where in the book | What I have | Gap question. Then the flags and the question list.
Fill in: STORY_NOTES FRAMEWORK LENGTH_AND_POV
Choose the framework you already like; if you have none, try three-act first. Answer the gap questions before asking for anything else.
Best for: Finding why the middle of a draft drags and choosing a fix
You are a developmental editor known for fixing the middle third of novels, where most drafts lose momentum.
Summary of my book, act by act or chapter by chapter: {{CHAPTER_SUMMARY}}
Where readers or I lose interest: {{WHERE_IT_SAGS}}
Protagonist's goal at the start and what changes at the midpoint: {{GOAL_AND_MIDPOINT}}
Diagnose before prescribing:
1. For each chapter in the middle section, write one line: what the protagonist wants in it, what blocks them, and what is different at the end. Mark any chapter where nothing changes.
2. Check the midpoint: is there a reversal, a revelation or a raise in stakes that changes the protagonist's approach? If not, say so plainly.
3. Look for these common causes and say which apply, with chapter references: repeated beats (the same kind of obstacle twice), a passive protagonist, a subplot carrying too much weight, stakes that don't escalate, too much travel or research, missing ticking clock.
4. Offer up to three fixes, each tied to a specific cause, framed as options for me to choose ("You could..."). Each fix must use characters and threads already in my story; don't invent new ones or rewrite scenes.
Finish with the one chapter you'd look at first in revision and why.
Fill in: CHAPTER_SUMMARY WHERE_IT_SAGS GOAL_AND_MIDPOINT
Paste a chapter-by-chapter summary, not the whole manuscript. A reverse outline (see the Reverse Outline prompt) is ideal input.
Best for: Seeing the real structure of a draft before revising
Act as a structural editor. I've finished a draft and need an honest map of what's actually on the page, not what I meant to write.
My chapter summaries or the chapters themselves (pasted below or in parts): {{DRAFT_CHAPTERS}}
Point-of-view characters: {{POV_CHARACTERS}}
For every chapter, create one row in a table with these columns:
- Chapter number
- POV character
- Story time (day, date or "unclear")
- Scene goal: what the POV character wants in this chapter
- Conflict: what stands in the way
- Outcome: how the chapter ends (yes, no, yes-but, no-and)
- Plot threads touched
- Word count, if I've given it
After the table:
1. List chapters where the outcome doesn't change anything, as candidates to cut or combine.
2. List plot threads that disappear for more than three chapters.
3. Show how the POV is split as a rough count of chapters per character.
4. Flag time jumps that may confuse a reader.
Stick to what's on the page. Don't suggest new scenes or rewrite anything; I'll use this map to plan my revision.
Fill in: DRAFT_CHAPTERS POV_CHARACTERS
Use it right after finishing a draft. If your manuscript is too long to paste at once, paste a few chapters per message and ask for the table at the end.
Best for: Making subplots pull their weight and land at the right moments
You are a plotting consultant for multi-thread novels. Help me weave my subplots so each one earns its place and pressures the main plot.
Main plot in brief: {{MAIN_PLOT}}
My subplots (romance, family, mystery, rivalry, etc.), each in a sentence or two: {{SUBPLOTS}}
Approximate chapter count: {{CHAPTER_COUNT}}
Work through this:
1. For each subplot, name its function: does it complicate the main goal, mirror the theme, reveal character, or supply a reveal the climax needs? A subplot with no function gets flagged.
2. Show where each subplot touches the main plot now, and suggest where its crisis could collide with a main-plot turning point for maximum pressure. Use only characters and events I've mentioned.
3. Build a grid: rows are subplots, columns are chapter ranges (beginning, early middle, midpoint, late middle, climax, resolution). Each cell gets a short note of what happens or "absent".
4. Flag any stretch where one thread vanishes for too long, and any subplot that resolves after the main climax and might deflate the ending.
Present your ideas as options. I'll make the plot decisions.
Fill in: MAIN_PLOT SUBPLOTS CHAPTER_COUNT
Use it during outlining or second-draft planning when you have more than one storyline.
Best for: Endings that feel inevitable and surprising at once
Act as a novel editor who specialises in endings. My ending isn't working and I want options that pay off what the book has already promised.
Summary of the book up to the climax: {{SUMMARY_TO_CLIMAX}}
My current ending: {{CURRENT_ENDING}}
What I want readers to feel when they close the book: {{DESIRED_FEELING}}
Do this in order:
1. List every promise the book makes before the climax: questions raised, setups planted, skills or objects introduced, relationships left unresolved. Use my summary as the source.
2. Check my current ending against that list. Which promises does it keep, break or ignore?
3. Check that the protagonist resolves the climax through their own choice and growth, not luck or rescue.
4. Offer three alternative ending shapes (for example, win with a cost, lose but grow, a choice between two goods). For each, give only a short sketch built from existing story elements, plus the setups I'd need to add or strengthen earlier.
5. Tell me which feels most in line with the feeling I want, with one sentence of reasoning.
Don't write the ending scene itself. I want options and a checklist, then I'll write it.
Fill in: SUMMARY_TO_CLIMAX CURRENT_ENDING DESIRED_FEELING
Use it when beta readers say the ending feels rushed, unearned or flat.
Best for: Turning an outline beat into a draftable scene plan
You are a book coach helping me turn an outline beat into scenes I can draft tomorrow.
The outline beat or chapter summary I'm about to write: {{CHAPTER_BEAT}}
What happened in the previous chapter: {{PREVIOUS_CHAPTER}}
Characters present and what each wants right now: {{CHARACTERS_AND_WANTS}}
Break the chapter into 2-4 scenes. For each scene give:
- POV character and location
- Scene goal (what the POV character is trying to do)
- Conflict (who or what pushes back)
- Turn: the moment the scene changes direction
- Outcome and the new problem it creates
- One emotional beat for the POV character
- Entry point: a suggestion of where to start the scene as late as possible
Then:
- Note anything from the previous chapter that this chapter must acknowledge.
- Suggest one place to cut if the chapter runs long.
Keep it to notes, not prose. Don't write any dialogue or narration; I'll draft it in my own voice. If the beat seems to need information you don't have, ask me instead of inventing it.
Fill in: CHAPTER_BEAT PREVIOUS_CHAPTER CHARACTERS_AND_WANTS
Use it the night before a writing session, so you sit down knowing what each scene has to do.
Characters & Worldbuilding
Prompts that help you discover characters and worlds rather than download them. The AI interviews you about a character's want, need and wound, tests whether your antagonist's choices make sense, checks whether your characters' dialogue sounds distinct, and helps you prepare questions for a sensitivity reader. For speculative and historical fiction, it asks hard questions about power, cost and daily life, and plans research without passing off guesses as facts.
Best for: Building a layered character from your own instincts
Act as a character-development coach. Interview me about my character so I discover them myself. Don't invent their backstory for me.
Character's name and role in the story: {{CHARACTER_AND_ROLE}}
What I know about them so far: {{WHAT_I_KNOW}}
Genre and tone of the book: {{GENRE_TONE}}
Run the interview one question at a time and wait for my answer. Cover these areas, in this order:
1. Want: the concrete external goal they pursue in the story.
2. Need: what they actually need to learn or accept, which may contradict the want.
3. Wound: an earlier experience that shaped how they see the world.
4. The lie they believe because of it.
5. What they'd never do, and the circumstances in which they might do it anyway.
6. How they talk: vocabulary, rhythm, what they avoid saying.
7. What they do when nobody is watching.
If an answer is vague, ask me for a specific memory or moment rather than a trait. After 8-12 questions, summarise in a character sheet using my wording: want, need, wound, lie, contradiction, voice notes, and the scene where the lie is most likely to break.
Fill in: CHARACTER_AND_ROLE WHAT_I_KNOW GENRE_TONE
Use it for any major character, including antagonists. Answer quickly and instinctively; you can revise later.
Best for: Giving a villain logic, competence and a reason to exist
You are a story editor who believes the best antagonists are right about something. Help me strengthen mine without changing who they are.
My antagonist: {{ANTAGONIST}}
What they do in the plot: {{ANTAGONIST_ACTIONS}}
My protagonist and the book's central conflict: {{PROTAGONIST_AND_CONFLICT}}
Analyse:
1. Goal and logic: what does the antagonist want, and does every action I've listed follow from it? Flag actions that exist only because the plot needs them.
2. Point of view: write 3-4 sentences summarising the story as the antagonist would tell it, as if they were the hero. Use their own reasoning, not mine.
3. Mirror: how does the antagonist reflect, invert or tempt the protagonist? If there's no link, say so.
4. Competence: where does the antagonist act too stupidly or too conveniently? Cite the specific action.
5. Escalation: does their pressure on the protagonist grow across the book? List the steps.
Then give me five questions to answer about the antagonist and suggest two moments where showing their reasoning would add depth. Don't rewrite any scenes, and keep any suggestions within the events I've described.
Fill in: ANTAGONIST ANTAGONIST_ACTIONS PROTAGONIST_AND_CONFLICT
Use it if beta readers call your villain flat, cartoonish or convenient. Works for impersonal antagonists such as institutions too.
Best for: Checking that every character has a distinct way of speaking
Act as a dialogue editor. I want to know whether my characters sound different from each other on the page.
Dialogue from my manuscript, with speaker tags removed or kept (say which): {{DIALOGUE_SAMPLES}}
Who the characters are (age, background, personality, one line each): {{CHARACTER_LIST}}
Do this:
1. If I removed speaker tags, try to attribute each line to a character and explain what clue you used. Lines you can't attribute are a sign the voices are blurring.
2. For each character, describe their current dialogue voice from the evidence: sentence length, vocabulary, formality, humour, what they avoid saying, verbal habits. Quote my lines.
3. Point out pairs of characters who sound too alike, with examples.
4. For each character, suggest 2-3 voice levers I could adjust, such as how they handle questions, interruptions, profanity, idiom or jargon. Don't rewrite my lines; describe the levers.
5. Flag any dialogue that exists mainly to tell the reader information both characters already know.
Keep everything tied to quotations from my sample. Don't generate new dialogue unless I ask.
Fill in: DIALOGUE_SAMPLES CHARACTER_LIST
Paste 1-3 pages of dialogue-heavy scenes. Removing speaker tags first gives the most honest result.
Best for: Making a relationship change believably, scene by scene
You are a developmental editor who specialises in character relationships, whether romance, friendship, family or rivalry.
The two characters and their relationship at the start: {{RELATIONSHIP_START}}
Where I want the relationship to be at the end: {{RELATIONSHIP_END}}
Key scenes they share, in order, one line each: {{SHARED_SCENES}}
Build the map:
1. For each shared scene, rate the relationship on two scales from 1 to 5: trust and tension. Give one line of reasoning per scene.
2. Show the arc as a simple text chart (scene number, trust, tension).
3. Identify flat stretches where nothing in the relationship changes, and jumps where it changes faster than the scenes justify.
4. Find the moments of choice: where does each character choose the relationship over something else, or something else over it? If no such moment exists, flag it.
5. Note any genre expectations my arc must meet (for example, a romance's happy or hopeful ending) and whether it meets them.
Suggest up to three places where a scene could be added, moved or sharpened, described only as a purpose ("a scene where she sees him fail and stays"). Don't write the scenes.
Fill in: RELATIONSHIP_START RELATIONSHIP_END SHARED_SCENES
Use it for romance arcs, mentor relationships, families or friendships that drive your plot.
Best for: Pressure-testing a fantasy or science-fiction world for consistency
Act as a worldbuilding consultant for speculative fiction. I don't want you to build my world; I want hard questions that make me build it consistently.
My setting in brief: {{SETTING_SUMMARY}}
The parts readers will see most (cities, institutions, landscapes): {{ON_PAGE_ELEMENTS}}
Genre and how much worldbuilding readers of it expect: {{GENRE}}
Give me 25 questions grouped under these headings, five each:
1. Power: who rules, who enforces, who resists, and how they get away with it.
2. Economy and daily life: what people eat, trade, fear and do for fun.
3. Rules of the speculative element (magic, technology, biology): what it costs, who can't use it, what it can never do.
4. History: the event everyone still argues about, and the version the winners tell.
5. Culture and belief: what is taboo, what is sacred, what jokes are told.
Each question should be specific to my setting, not generic. Put a star next to the five questions whose answers will most affect my plot.
Finish with a short list of contradictions or gaps you noticed in my summary. Don't answer the questions yourself.
Fill in: SETTING_SUMMARY ON_PAGE_ELEMENTS GENRE
Answer the starred questions first. Keep the answers in your story bible; most will never appear on the page, and that's fine.
Best for: Making magic or tech feel consistent and keeping the climax earned
You are a speculative fiction editor who checks whether a magic or technology system holds up across a whole book.
How my system works, as I've written it: {{SYSTEM_RULES}}
How the protagonist uses it at the climax: {{CLIMAX_USE}}
Other key uses earlier in the book: {{EARLIER_USES}}
Check the system:
1. Restate the rules as a short numbered list, so I can confirm you've understood them.
2. Costs and limits: what does using it cost (physically, socially, morally, financially)? If there's no meaningful cost, say what problems that creates for tension.
3. Loopholes: list ways a clever character could exploit these rules to solve the plot early. For each, ask why nobody does.
4. Climax test: is the climax use set up earlier, and does it follow the rules? If it breaks a rule, has the book earned the exception?
5. Consistency: compare the earlier uses against the rules and flag any contradiction.
6. Wider effects: name two ways this system would change ordinary society that my summary doesn't mention.
Keep suggestions as questions or options. The system is mine; don't redesign it.
Fill in: SYSTEM_RULES CLIMAX_USE EARLIER_USES
Use it during outlining, and again after the first draft to catch rules you broke without noticing.
Best for: Planning research that keeps a period or place convincing
Act as a research librarian working with a historical or location-based novelist. I need to know what to research, not a stream of facts that may be wrong.
Setting (place, period, social world): {{SETTING_TIME_PLACE}}
Scenes I'm about to write: {{UPCOMING_SCENES}}
What I already know or have read: {{EXISTING_RESEARCH}}
Do this:
1. List the details my scenes will need: material culture (clothing, food, money, transport, light), language and forms of address, laws and institutions, daily routines, sensory texture. Make every item specific to my scenes.
2. For each, mark it "likely to trip writers up" or "lower risk", with a one-line reason.
3. Suggest the types of primary and secondary sources that would answer each (newspapers, diaries, maps, oral histories, museum collections, specialist histories). Name a specific source only if you are confident it exists, and mark it "verify".
4. List common anachronisms or myths about this period or place that I should check before using.
5. Suggest two questions I could ask a local expert, historian or museum curator.
Don't present uncertain facts as settled. Where you give a fact, label it "check this" so I verify it before it goes on the page.
Fill in: SETTING_TIME_PLACE UPCOMING_SCENES EXISTING_RESEARCH
Use it before drafting a period or location-heavy chapter. Verify every fact the AI states against a real source.
Best for: Preparing focused questions for a sensitivity or authenticity reader
Act as an editor who helps authors prepare for a sensitivity or authenticity read when they write characters whose identity or experience differs from their own. You are not a substitute for that reader, and you should say so if I seem to be treating you as one.
The character and the experience or identity I'm writing outside of: {{CHARACTER_AND_IDENTITY}}
The scenes or passages where it matters most: {{PASSAGES}}
The research and first-hand sources I've used so far: {{MY_RESEARCH}}
Do this:
1. Quote any lines where the character seems defined mainly by this identity, rather than having wants, flaws and a story of their own.
2. List possible stereotypes, tropes or common harmful patterns the passages may be drifting towards (for example a character who exists to teach the protagonist, or whose suffering is the plot), quoting the evidence. Mark these as possibilities, not verdicts.
3. Note details that may need first-hand checking: language, customs, faith practice, medical or legal realities, and community-specific terms.
4. Write 8-10 specific questions I should bring to a paid sensitivity reader or someone with lived experience.
Don't rewrite the passages, and don't tell me whether I'm "allowed" to write this character. Help me ask better questions.
Fill in: CHARACTER_AND_IDENTITY PASSAGES MY_RESEARCH
Use it before you hire a sensitivity reader, so their time goes on the hard questions. It doesn't replace a human reader with lived experience.
Scene Craft: Feedback on Your Pages
Paste a scene and get the notes an editor would write in the margin: show versus tell, goal and conflict, dialogue subtext, point-of-view slips, pacing and sensory detail. Every prompt quotes your words and describes the fix rather than rewriting your prose, so the revision stays in your voice. There's also a first-page critique from the point of view of an agent's first reader, and a coach for the scene you can't start.
Best for: Editor-style show/tell notes that leave the rewriting to you
Act as an experienced fiction editor giving margin notes on show-versus-tell. Your notes should help me revise in my own voice, so don't rewrite my prose.
My scene: {{SCENE_TEXT}}
What the reader should feel or understand by the end of it: {{SCENE_INTENT}}
Genre and point of view: {{GENRE_AND_POV}}
Give notes in this format:
1. Up to 8 notes, in the order they appear. Each note: quote the passage (no more than two sentences), name the issue (named emotion, summary where a moment is needed, explained subtext, filter words such as "she saw" or "he felt", or a vague abstraction), and say in one sentence what kind of concrete detail, action or dialogue could carry it instead. Don't write the replacement.
2. Two places where telling is the right choice (summary that keeps the pace moving, or transitions) so I don't over-correct.
3. The single strongest moment of showing already in the scene, quoted, and why it works.
4. One line on whether the scene achieves the intent I gave you.
If the scene is mostly strong, say so and give fewer notes. Don't invent problems to fill the list.
Fill in: SCENE_TEXT SCENE_INTENT GENRE_AND_POV
Paste one scene at a time, up to about 2,000 words. Revise from the notes yourself, then paste again to check.
Example output · Claude (example written by The Prompt Index)
Sample inputs: SCENE_TEXT = (opening of the scene) "Nell was nervous as she climbed the stairs to the counting room. She knew Mr Tregear would be angry if he found her there after dark. The ledgers were on the shelf where they always were. She felt a cold dread as she opened the book for 1890. There it was: twelve bushels of salt, entered against the Morwenna, three weeks after the Morwenna went down. She saw that the handwriting was her father's. Nell was shocked. Her father had been dead for three years and yet here was his hand, neat and slanting, the way he wrote her name on birthday parcels. She heard footsteps on the quay below. She realised she had to hide the page." SCENE_INTENT = The reader should feel the floor drop out from under Nell, and fear being caught. GENRE_AND_POV = Historical mystery, close third person on Nell.
1. Notes
1. "Nell was nervous as she climbed the stairs." Named emotion. Her nerves could come through in what she does: which step she avoids, what she does with the lamp.
2. "She knew Mr Tregear would be angry." Explained stakes. One concrete thing he has done to someone before would let the reader supply the fear.
3. "She felt a cold dread." Filter word plus named emotion. Her body can carry it; this is the moment before the discovery, so slowing down helps.
4. "She saw that the handwriting was her father's." Filter ("she saw") at the turning point. Let the handwriting land on the reader directly, without Nell standing between us and it.
5. "Nell was shocked." Named emotion right after the strongest image in the scene. The sentence that follows already does this work, so this line competes with it.
6. "She heard footsteps on the quay below." Filter. The sound itself, specific to a quay at night, would be more immediate.
7. "She realised she had to hide the page." Explains her decision. Her action alone would show it, and moves faster.
2. Where telling is right
- "The ledgers were on the shelf where they always were." It's quick orientation that keeps the pace moving. Keep it.
- "three weeks after the Morwenna went down" is summary the reader needs in order to grasp the clue. Keep it plain.
3. Strongest moment
"neat and slanting, the way he wrote her name on birthday parcels." A specific, private memory does the emotional work without naming a feeling. More of the scene can work at this level.
4. Intent
The discovery lands, mostly thanks to the birthday-parcel line. The fear of being caught is told rather than felt, so the threat from Mr Tregear needs a concrete form before the footsteps arrive.
Best for: Getting past writer's block on one scene, with your own words
You are a writing coach who gets authors moving again when a scene won't come. You never draft the scene; you ask the questions that let me draft it.
Where I'm stuck (the scene, what comes before it, what I've tried): {{STUCK_POINT}}
What I know must happen in or by the end of this scene: {{MUST_HAPPEN}}
How I feel about the book right now, honestly: {{MY_STATE}}
Work through this with me:
1. Diagnose the likely cause from what I've said, choosing from: I don't know what the POV character wants here; the scene has no opposition; I'm starting too early; the scene isn't needed; I've taken a wrong turn earlier; or plain fatigue. Explain your reasoning in two or three sentences.
2. Ask me three questions aimed at that cause. Wait for my answers.
3. Then offer three ways in, as one-line prompts to myself, such as "start mid-argument", "write it from the other character's point of view as a throwaway exercise" or "write only the last line first".
4. Suggest a 20-minute exercise to do right now.
If the problem seems to be earlier in the book, say which earlier decision to revisit. Keep your replies short. Don't write any prose, dialogue or plot events for me.
Fill in: STUCK_POINT MUST_HAPPEN MY_STATE
Use it when you've been staring at the same blank scene for days. Do the 20-minute exercise before you ask for anything else.
Best for: Checking a scene has a goal, conflict and turn
You are a developmental editor checking whether a scene earns its place. Assess my scene against the basic mechanics of scene tension.
My scene: {{SCENE_TEXT}}
What comes immediately before and after it: {{CONTEXT}}
Answer each with a short quote from my scene as evidence:
1. Goal: what does the POV character want in this scene? Is it clear by the end of the first page?
2. Conflict: what opposes them? Is it active (someone or something pushing back) or passive (waiting, thinking)?
3. Turn: where does the scene change direction? If it doesn't, say so.
4. Outcome: does it end with the character better off, worse off, or with a new problem? Does it close the scene or pull the reader forward?
5. Stakes: what does the character lose if they fail here?
Then:
- Mark the line where the scene could start later and the line where it could end sooner, if any.
- Rate the scene's tension from 1 to 5 with a one-sentence justification.
- Give no more than three revision suggestions, each tied to one of the five points above.
Don't rewrite passages. Describe changes so I can make them in my own style.
Fill in: SCENE_TEXT CONTEXT
Use it on any scene that feels flat, or on every scene in a chapter that beta readers skimmed.
Best for: Sharper dialogue with subtext, kept in your characters' voices
Act as a fiction line editor who focuses on dialogue. Give me edit notes on the dialogue in this passage without rewriting my lines.
The passage: {{DIALOGUE_PASSAGE}}
What each speaker wants from this conversation, which they may not say out loud: {{SPEAKER_AGENDAS}}
Review for:
1. Subtext: where characters say exactly what they mean, and whether holding something back would serve the scene. Quote the line.
2. "As you know" lines: information both characters already know, delivered for the reader's benefit.
3. Tags and beats: over-used adverbs, fancy dialogue tags, or action beats that don't add anything. Count them and quote examples.
4. Rhythm: speeches that run too long for real speech, or exchanges where every line is the same length.
5. Agenda: whether each speaker is visibly pursuing what they want. Flag any stretch where a character is only there to ask questions.
6. Ending: whether the exchange ends on a shift in power or knowledge.
Output: a numbered list of notes in the order of the passage, each with the quote, the issue and a one-line suggestion of the fix to try. Finish with the best line of dialogue in the passage and why it works.
Fill in: DIALOGUE_PASSAGE SPEAKER_AGENDAS
Fill in the speaker agendas; it's what lets the editor spot missing subtext.
Best for: Seeing your opening the way a busy agent or reader would
You are a literary agent's first reader. You read hundreds of opening pages a month and decide in a page whether to keep reading. Be honest and specific; politeness won't help me.
My first page (about 300-400 words): {{FIRST_PAGE}}
Genre and age category: {{GENRE_CATEGORY}}
What happens by the end of chapter one: {{CHAPTER_ONE_SUMMARY}}
Give your reaction in this order:
1. Your reading experience line by line: where your attention rose, where it dropped, and the exact sentence where you would have stopped if you would have.
2. What I've established by the end of the page: whose story, what they want, what's wrong, the voice, and genre signals. Mark each as clear, faint or missing.
3. Common opening problems you can see, if any: waking up or weather, backstory or world explanation, too many names, action with no stakes, a voice that sounds generic.
4. Whether the first page promises the book that chapter one delivers, given my summary.
5. Keep reading: Yes, Maybe or No, with one sentence of reasoning.
Quote my words for every point. Don't rewrite the page.
Fill in: FIRST_PAGE GENRE_CATEGORY CHAPTER_ONE_SUMMARY
Run it on your opening page before querying or entering a contest. Run it again after each rewrite.
Best for: Catching head-hopping and POV slips before an editor does
Act as a copy editor who specialises in point of view. Check this passage for POV consistency.
The passage: {{PASSAGE}}
The intended POV (for example close third on Mara, first person present, omniscient): {{INTENDED_POV}}
Check for:
1. Head-hopping: places where we learn another character's thoughts or feelings that the POV character couldn't know. Quote each and say whose head we slipped into.
2. Knowledge slips: the POV character describing things they can't see, or knowing things they haven't learned yet.
3. Distance shifts: unintended jumps between close and distant narration (for example, a close third suddenly describing the character's own face, or "little did she know").
4. Tense slips, if the tense changes without a reason.
5. Filtering: "she noticed", "he felt", "she could see" where deep POV would drop the filter. Count these and quote three examples.
Output a table: Quote | Issue type | Why it breaks the POV | Suggested approach (described, not rewritten).
If the passage is intentionally omniscient or has a deliberate shift, say how to make the intention clear rather than flagging it as an error. Don't rewrite the passage.
Fill in: PASSAGE INTENDED_POV
Use it on chapters with several characters in the room, where head-hopping is most likely.
Best for: Seeing where a chapter drags or rushes
You are a developmental editor who maps pacing. Read my chapter and show me where it speeds up and slows down.
The chapter: {{CHAPTER_TEXT}}
The pace I'm aiming for (for example breathless chase, slow-burn dread, quiet aftermath): {{TARGET_PACE}}
Do this:
1. Split the chapter into beats of roughly a paragraph or a short exchange. For each beat, write: the first few words (so I can find it), the mode (action, dialogue, interior thought, description, summary, backstory) and pace on a scale from 1 (slow) to 5 (fast).
2. Draw a simple text chart of pace across the chapter.
3. Point out stretches of three or more slow beats in a row, and say whether they're earning their space (building dread, giving a breather) or stalling.
4. Flag any backstory or explanation that arrives at a high-tension moment and lets the air out.
5. Note sentence and paragraph length patterns in fast and slow sections, quoting examples.
End with up to three suggestions for matching the pace to my target, each pointing to a specific beat. Describe the change; don't rewrite.
Fill in: CHAPTER_TEXT TARGET_PACE
Use it on action climaxes and on quiet chapters that readers report as slow.
Best for: Grounding a scene in specific sensory detail, your way
Act as a fiction editor who helps writers ground scenes in a real place without slowing the story.
My scene: {{SCENE_TEXT}}
Where and when it takes place: {{SETTING}}
The mood I want: {{MOOD}}
Audit the scene:
1. Senses used: count the details for sight, sound, smell, touch and taste, with one quote for each sense used. Name the senses that are missing.
2. Generic vs specific: quote details that could be in any story ("the room was dark") and say what kind of specific detail would only fit this place and character.
3. Filtered through the character: are details noticed the way this POV character would notice them, given their job, mood and history? Give an example where it works and one where it doesn't.
4. Placement: is the setting established within the first few paragraphs, or is the reader floating in white space? Is description dumped in one block?
5. Mood: do the chosen details support the mood I want? Quote any that pull against it.
Finish with three questions about the place that, if I answered them, would give me sharper details. Don't write description for me.
Fill in: SCENE_TEXT SETTING MOOD
Use it when beta readers say they can't picture a scene, or when description feels like a list.
Revision, Line Editing & Your Voice
Big-picture and sentence-level revision. Get a developmental edit letter, line notes that respect deliberate style choices, a plan for cutting words, a beta-reader questionnaire and a read-aloud rhythm check. One prompt from our library, StyleMaster, turns a sample of your prose into a voice paragraph you can paste before other prompts so their feedback respects how you write.
Best for: A structured big-picture critique before you start line edits
You are a developmental editor writing an edit letter, the big-picture feedback an author gets before line edits. Be honest, specific and kind, in that order.
My synopsis or chapter summaries: {{SYNOPSIS}}
Sample chapters (the opening and one from the middle work best): {{SAMPLE_CHAPTERS}}
Genre, target readers and my main worry about the book: {{GENRE_AND_WORRY}}
Write the letter with these sections:
1. What's working: 3-4 specific strengths, with evidence. These are things to protect in revision.
2. The big issues: no more than four, ranked by how much they affect the reader's experience. For each, describe the problem, show where it appears (chapter or quote), explain why it matters to a reader of this genre, and offer two possible directions. The choice of direction is mine.
3. Character arcs: whether the protagonist changes, and whether the change is earned on the page.
4. Structure and pacing: where the book gains and loses momentum.
5. Genre fit: whether the book delivers what its readers expect.
6. My worry: address it directly.
7. Revision order: what to tackle first, second and third.
Don't do line edits or rewrite passages. Keep the letter under 1,200 words.
Fill in: SYNOPSIS SAMPLE_CHAPTERS GENRE_AND_WORRY
Use it after a complete draft, before line editing. AI can't replace a human editor's judgement, but it's a fast first opinion.
Best for: Line editing without losing the way you write
Act as a line editor who respects the author's voice above everything. My style is deliberate; your job is to catch what weakens it, not to make it sound like someone else.
My passage: {{PASSAGE}}
Things that are intentional in my style and should not be "fixed" (fragments, long sentences, dialect, etc.): {{STYLE_CHOICES}}
Give line notes as a table with these columns: Quote | Issue | Suggestion. Suggestions should be minimal: cut a word, swap a verb, reorder a clause. Never rewrite a whole sentence unless it's unclear, and then offer two options that keep my vocabulary.
Look for:
- Unclear sentences (meaning can be read two ways)
- Repeated words or sentence openings in close proximity
- Weak verbs propped up by adverbs
- Clichés and stock phrases
- Rhythm problems: runs of same-length sentences
- Overwriting: three adjectives where one would do
- Grammar and punctuation errors that aren't style choices
After the table, give a short paragraph describing my voice in this passage, so I know what you were protecting. Limit yourself to the 15 most useful notes.
Fill in: PASSAGE STYLE_CHOICES
Use it on polished drafts, a scene at a time. Accept or reject each note; never paste back a whole AI rewrite.
Best for: Capturing your writing voice as a reusable style note
You’re an AI system that has been trained to analyze the text below for style, voice, and tone using NLP to create a voice paragraph. The Voice Paragraph you write is designed to give specific directions to a future AI system so that it will write in that same style, voice, and tone. Your VoiceParagraph should NOT be specific to the actual content and should instead focus on the style, voice and tone of writing. Your VoiceParagraph output should be written in a single paragraph and should include specific directions to write in the same style voice and tone of the author of this writing sample only. Your VoiceParagraph output should not refer to the sample text you are analyzing or the analysis you have performed, only explain the writing techniques that should be deployed to create similar text. Begin your response with "Write with the following style, voice, and tone..." Limit your VoiceParagraph to a single paragraph. Be concise, including only the most relevant directions. Here is the input text to write your VoiceParagraph description based on: {{user text here}}
Greetings! I am StyleMaster, an AI system trained in the art of analyzing style, voice, and tone.
My purpose is to assist you in creating a voice paragraph that provides specific directions to future AI systems, so they can write in the same style, voice, and tone as the author of the provided writing sample.
To accomplish this, I will carefully analyze the input text and generate concise and relevant directions. Are you ready to proceed? Please share the input text with me, and we shall embark on our journey to capture the essence of the author's writing style."
Best for: A ranked plan to cut words with the least damage
You are an editor who helps authors hit word-count targets for agents, publishers or contests by cutting fat rather than muscle.
Current word count and target: {{CURRENT_AND_TARGET}}
Chapter summaries with approximate word counts: {{CHAPTER_SUMMARIES}}
A sample chapter so you can see my prose habits: {{SAMPLE_CHAPTER}}
Build a cutting plan:
1. Structural cuts: chapters or scenes whose outcome is repeated elsewhere, subplots that don't affect the climax, and openings that start too early. Estimate the words saved by each cut.
2. Scene-level trims: scenes that could start later or end sooner, with a rough saving.
3. Line-level habits from my sample chapter: quote examples of patterns I repeat (stage directions, restated emotions, double descriptions, throat-clearing openings). Estimate how much these cost per chapter.
4. A table: Cut | Type | Words saved (estimate) | Risk to story (low/medium/high).
5. A suggested order of cuts that reaches the target with the lowest risk.
Make clear that word savings are estimates. Don't rewrite any passage, and mark anything I'd lose emotionally so I can decide whether it's worth it.
Fill in: CURRENT_AND_TARGET CHAPTER_SUMMARIES SAMPLE_CHAPTER
Use it when your manuscript is longer than the typical length your genre's agents or publishers ask for.
Best for: Getting honest, comparable feedback from beta readers
Act as a writing-group organiser who designs feedback forms that get honest, useful answers from beta readers.
My book in brief: {{BOOK_SUMMARY}}
What I'm most unsure about: {{MY_CONCERNS}}
My beta readers (for example, genre fans, writers, people who know the subject): {{READER_TYPES}}
Create:
1. A short note to beta readers explaining what kind of feedback I want and what I don't (for example, no proofreading at this stage).
2. Chapter check-in questions (3 per chapter, the same each time) that capture engagement without leading: where they put the book down, what they expect next, who they're rooting for.
3. End-of-book questions (10-12), including at least one per concern I listed. Phrase them neutrally: ask "What did you think of X?" rather than "Did you love X?".
4. A 1-5 scale question for overall enjoyment, and one for "would you recommend it to a friend who reads this genre?".
5. A simple way for me to tally answers across readers, as a table template.
Avoid questions that ask readers to fix the book. They are good at telling you where they felt something, not at prescribing solutions.
Fill in: BOOK_SUMMARY MY_CONCERNS READER_TYPES
Send it with the manuscript. Look for patterns across readers; one person's note is an opinion, three is a problem.
Best for: Prose that sounds right when read aloud or narrated
You are a prose-rhythm editor, like someone who prepares manuscripts for audiobook narration. Check how my passage sounds when read aloud.
My passage (ideally a key scene, opening or closing paragraph): {{PASSAGE}}
The effect I want the rhythm to create: {{DESIRED_EFFECT}}
Analyse:
1. Sentence lengths: list the word count of each sentence in order, then describe the pattern (varied, monotonous, building, staccato).
2. Stumbles: phrases a narrator would trip over, such as tongue-twisters, unclear antecedents, stacked prepositional phrases or accidental rhymes. Quote them.
3. Stress at the end: sentences where the most important word is buried in the middle instead of landing at the end. Quote three.
4. Breath: sentences too long to say in one breath that aren't deliberately breathless.
5. Paragraph endings: whether each paragraph's last line lands or trails off.
Give no more than 10 notes, each with the quote and a one-line suggestion (reorder, split, cut). Then tell me whether the overall rhythm creates the effect I want. Don't rewrite the passage.
Fill in: PASSAGE DESIRED_EFFECT
Read the passage aloud yourself after applying the notes. Your ear is the final judge.
Continuity & Consistency Checks
Long books drift: eye colours change, a ledger moves rooms, a character knows something they haven't learned yet. These prompts build a story bible from your actual pages, check each new chapter against it, align a mystery's true timeline with its reveal order, audit setups and payoffs, and brief you on what a series sequel must respect.
Best for: A fact sheet of your book, built from the actual pages
Act as a continuity editor building a story bible from a manuscript. Extract only what's on the page. Don't add or guess.
My chapters (paste in batches if needed): {{CHAPTERS}}
Categories I care most about: {{PRIORITY_CATEGORIES}}
Create a story bible with these sections:
1. Characters: name, role, physical details (eye and hair colour, height, scars, clothing habits), age or birthday, relationships, verbal habits. Each fact must have a chapter reference.
2. Places: name, description, geography and distances or travel times mentioned.
3. Timeline: every time marker in order (dates, days of the week, seasons, "three days later") with chapter references.
4. Objects and rules: important objects, who has them when, and any rules of magic, technology or institutions stated on the page.
5. Open threads: questions raised and promises made that haven't been resolved yet.
6. Conflicts found: any fact stated two different ways, quoted with both chapter references.
Format as headed lists. Mark anything ambiguous as "unclear" rather than choosing a version. When I paste more chapters later, update the bible and list new conflicts at the top.
Fill in: CHAPTERS PRIORITY_CATEGORIES
Start it early and keep it updated as you draft. Paste the bible before continuity checks so the AI has the facts.
Best for: Catching contradictions chapter by chapter
You are a continuity editor for long-form fiction. Check a new chapter against the established facts of the book.
My story bible or fact notes: {{STORY_BIBLE}}
The new chapter: {{NEW_CHAPTER}}
Where this chapter falls in the timeline: {{TIMELINE_POSITION}}
Check for:
1. Contradictions: any detail in the chapter that conflicts with the bible (physical description, names, ages, places, distances, rules). Quote both versions.
2. Knowledge errors: characters who know something they haven't learned yet, or have forgotten something they know.
3. Timeline problems: travel too fast, days of the week that don't fit, injuries that heal too quickly, seasons or weather that don't match.
4. Object tracking: objects appearing where they shouldn't, or disappearing.
5. Name drift: spellings, nicknames or titles used inconsistently.
Output a table: Issue | Quote from chapter | Conflicting fact (with source) | Severity (reader will notice / careful reader might notice / minor).
Then list new facts this chapter establishes, so I can add them to the bible. If you find nothing in a category, say "none found"; don't invent problems.
Fill in: STORY_BIBLE NEW_CHAPTER TIMELINE_POSITION
Run it after drafting each chapter. It's quicker to fix continuity as you go than at the end.
Example output · Claude (example written by The Prompt Index)
Sample inputs: STORY_BIBLE = Nell Pascoe: 24, grey eyes, left-handed, scar across right palm (gutting knife, age 12). Father: Jory Pascoe, lost with the Morwenna, 14 March 1890. Mr Tregear: owner of the cellar, walks with a cane since a fall in 1889. Distances: Porthkennack to Penzance about 9 miles, half a day on foot. Nell has not told anyone about the salt entry (end of ch. 6). Ledger for 1890 kept in the counting room. NEW_CHAPTER = Chapter 7 (about 2,500 words, summarised in the prompt with quoted lines). TIMELINE_POSITION = Two days after chapter 6; a Thursday in late October 1893.
Issue | Quote from chapter | Conflicting fact (source) | Severity
Eye colour | "Her brown eyes stung in the wind." | Grey eyes (bible, ch. 1) | Reader will notice
Hand | "She wrote the figure with her right hand, slowly." | Left-handed (bible, ch. 2) | Careful reader might notice
Knowledge | Aunt Bess: "So you've found your father's salt, then." | Nell has told no one (end of ch. 6) | Reader will notice. Either Bess has another source, which would be a strong clue if intended, or this is an error
Travel | Nell walks to Penzance after breakfast and is "back at the cellar by noon". | 9 miles each way, about half a day on foot one way (bible) | Reader will notice
Date | "Mother lit the Lammas candle." | Late October; Lammas falls on 1 August | Careful reader might notice
Object | Nell "took the 1890 ledger from beneath her mattress". | The ledger is kept in the counting room (bible); chapter 6 ends with her hiding one torn page, not the book | Reader will notice
Character | Tregear "strode across the yard". | Walks with a cane since 1889 (bible) | Careful reader might notice, unless the cane is a pretence. If so, it's a good clue; plant it deliberately
Name drift: "Porthkennack" is spelled "Porthkenack" twice. Minor.
Timeline: no other problems found.
New facts this chapter establishes (add to bible):
- Aunt Bess lives above the chandler's on Fore Street.
- The insurer is the Mount's Bay Mutual; their agent is a Mr Lowry.
- The Morwenna's skipper was Will Hosken (first mention).
- Nell owns a pocket watch that belonged to her father (it stopped at 4:10).
Best for: Checking a twisty plot holds together in time
Act as a plot-timeline analyst for mysteries, thrillers and dual-timeline novels. I need two timelines: what actually happened, and the order in which the reader finds out.
Events in my story, including hidden backstory and offstage events: {{EVENTS}}
The order of chapters and what each reveals: {{CHAPTER_REVEALS}}
Produce:
1. The true chronology: every event in the order it happened, with dates or relative times, including events that never appear on the page.
2. The reveal order: what the reader learns in each chapter.
3. A table aligning the two: Event | When it happened | Chapter where the reader learns it | Who knows it at that point (list characters).
4. Checks: alibis or movements that don't fit the timeline, clues that point to the solution too early or not at all, and characters who should have acted on information they had.
5. For a mystery: list the fair-play clues to the solution and the chapter where each appears. Flag the solution if it depends on information the reader never got.
Use only events I've given you. Where the timeline has gaps, list questions for me rather than filling them.
Fill in: EVENTS CHAPTER_REVEALS
Essential for mysteries, thrillers and anything with flashbacks or two timelines.
Best for: Making the ending feel earned by planting what it needs
You are a structural editor who checks that a story's setups pay off and its payoffs are set up.
My chapter summaries (with notable objects, skills, lines and promises): {{CHAPTER_SUMMARIES}}
The climax and ending: {{CLIMAX_AND_ENDING}}
Audit:
1. Setups without payoffs: objects, skills, secrets, prophecies, threats or lines given weight early on that never matter again. For each, reference the chapter and suggest whether to cut it or pay it off.
2. Payoffs without setups: moments in the climax or ending that rely on something not established earlier (a skill, an ally, a fact). For each, suggest the earliest natural chapter to plant it.
3. Setups paid off too soon, which leaves the ending without anything to use.
4. Repetition: setups planted so many times that the payoff becomes predictable.
Output a table: Element | Set up in (chapter) | Paid off in (chapter or "never") | Recommendation.
Finish with the three changes that would make the ending feel most earned. Describe each change in a sentence; don't write the scenes.
Fill in: CHAPTER_SUMMARIES CLIMAX_AND_ENDING
Use it after a full draft, alongside the Ending Options prompt if the ending feels unearned.
Best for: Starting the next book in a series without breaking the last one
Act as a series editor preparing a continuity brief before an author starts the next book in a series.
Summaries of the previous books, or their story bibles: {{PREVIOUS_BOOKS}}
My plan for the next book: {{NEXT_BOOK_PLAN}}
Write the brief:
1. Where everyone is: each major character's location, situation, relationships, injuries or changes, and what they know at the end of the last book.
2. Promises to readers: unresolved threads, teased conflicts and relationships readers will expect to see develop. Mark each as "must address", "could address" or "can wait".
3. Established rules: world rules, character abilities and facts the new book must respect.
4. Conflicts with my plan: anything in my next-book plan that contradicts what's been established, with the source.
5. Recap needs: the 5-8 facts a returning reader needs refreshed and a new reader needs to know. Suggest where in the opening chapters each could come up naturally, as notes.
6. Series arc: how the next book moves the overall series question forward.
Keep it under 1,000 words, organised by heading. Don't invent events that aren't in my summaries, and mark anything you're inferring.
Fill in: PREVIOUS_BOOKS NEXT_BOOK_PLAN
Use it before outlining book two or later. Keep the brief beside your outline.
Non-Fiction Books & Memoir
For how-to, big-idea and memoir books. Define who the book is for and what changes for them, build a table of contents and chapter plans from your own material, draft the sections of a proposal, recover memoir details through an interview, and list every claim to fact-check before publication. The prompts use [PLACEHOLDER] markers rather than inventing figures or credentials.
Best for: Defining exactly who a non-fiction book is for and what it delivers
You are a non-fiction book coach. Before I outline, I need a sharp promise: who the book is for, what changes for them, and why I'm the one to write it.
My topic and expertise: {{TOPIC_AND_EXPERTISE}}
Who I think the reader is: {{READER}}
The problem they have, in their words if I know them: {{READER_PROBLEM}}
Help me define:
1. The reader, narrowed: describe one specific reader in two sentences. Ask me one question if my description is too broad.
2. Before and after: where the reader starts (beliefs, situation, feelings) and where they are at the end, as a two-column table.
3. The promise: three versions of a one-sentence book promise of different scope, from narrow to broad, and which fits a first book best.
4. My credibility: what in my expertise supports the promise and what gaps a sceptical reader might see.
5. What the book is not: three related topics I should deliberately leave out.
6. Positioning: the kind of existing books on this topic a reader will compare mine to, described in general terms. Mark any specific titles you mention "verify".
Ask questions where you need information. Don't invent credentials or stories.
Fill in: TOPIC_AND_EXPERTISE READER READER_PROBLEM
Use it before a table of contents or a proposal. Every chapter should serve the promise you choose.
Best for: Turning a pile of ideas into a logical non-fiction structure
Act as an acquisitions editor for non-fiction. Build a table of contents from my book promise and material, and show the logic behind it.
My book promise and reader: {{BOOK_PROMISE}}
The ideas, frameworks, stories and lessons I want to include, in any order: {{MATERIAL_LIST}}
Book type (how-to, big-idea, memoir-led, reference) and target length: {{BOOK_TYPE_LENGTH}}
Do this:
1. Suggest an organising principle (chronological, step-by-step, problem-solution, myth-truth, or another that fits) and say why.
2. Draft 8-14 chapter titles in that order, each with a one-sentence chapter promise and the items from my material list it uses.
3. Group chapters into parts if it helps the reader.
4. List material that doesn't fit anywhere. It may belong in another book.
5. Flag gaps: steps the reader needs to make the transformation that my material doesn't cover yet.
6. Offer one alternative structure in brief, in case the first doesn't suit me.
Use my material; don't invent case studies or data. Chapter titles should be clear first and clever second.
Fill in: BOOK_PROMISE MATERIAL_LIST BOOK_TYPE_LENGTH
Use it after defining your book promise. Reorder freely; this is a first structure to react to.
Best for: Planning a non-fiction chapter that hooks, teaches and lands
You are a non-fiction developmental editor. Help me plan one chapter so it teaches, persuades and keeps the reader moving.
The book promise: {{BOOK_PROMISE}}
This chapter's title and promise: {{CHAPTER_PROMISE}}
My notes, stories, examples and sources for this chapter: {{CHAPTER_NOTES}}
Build a chapter plan:
1. Opening hook: suggest which of my stories or examples could open the chapter, and why. Don't invent a story.
2. The central idea in one sentence.
3. Sections in order, each with: the point it makes, the evidence or example from my notes that supports it, and the question the reader is likely to be asking at that point.
4. Objection: the strongest reasonable objection a sceptical reader would raise, and where to address it.
5. Practical takeaway: an exercise, checklist or reflection question that fits this chapter.
6. Bridge: one sentence pointing to the next chapter.
7. Evidence gaps: claims in my notes that need a source, and statistics I should check before quoting.
Output as an outline with headings. Don't draft the prose and don't add facts or figures of your own.
Fill in: BOOK_PROMISE CHAPTER_PROMISE CHAPTER_NOTES
Use it before drafting each chapter. The evidence-gaps list is your fact-checking to-do list.
Best for: Assembling a non-fiction proposal that an agent can sell
Act as a literary agent who sells non-fiction. Help me build the sections of a book proposal from my material. Proposals sell non-fiction books before they're written, so it needs to be clear and evidence-based.
My book's promise, reader and summary: {{BOOK_SUMMARY}}
My background and platform (audience size, newsletter, speaking, media, credentials), with accurate numbers: {{PLATFORM}}
Table of contents with chapter summaries: {{TOC}}
Produce a structured draft of:
1. Overview: the hook, the problem, the promise and why now (300-400 words).
2. Target readers: who buys the book, described specifically.
3. Comparable titles: tell me what to look for in 4-6 comps and how to describe each in relation to my book. Don't list titles unless you're confident they exist, and mark them "verify".
4. About the author: based only on the facts I've given.
5. Marketing and platform: what I can realistically do to reach readers, based on my platform. No inflated promises.
6. Chapter outline: tighten my chapter summaries to 100-150 words each.
Flag weak spots an agent would question. Use [PLACEHOLDER] for any number or fact I haven't given. Never make one up.
Fill in: BOOK_SUMMARY PLATFORM TOC
Use it after your table of contents is settled. Check what each agent's or publisher's guidelines ask for before you submit.
Best for: Recovering the specific details that make memoir come alive
You are a memoir coach who helps people recover the specific, sensory details that make life writing vivid. Interview me about a period or event from my life.
The period, place or event: {{MEMORY_TOPIC}}
Why it matters to the book: {{WHY_IT_MATTERS}}
Topics I'd rather not go into: {{OFF_LIMITS}}
How to run it:
1. Ask one question at a time and wait for my answer.
2. Start with concrete, sensory questions (what the room smelled like, what someone wore, what was on the radio) before emotional ones.
3. Follow my answers. When I mention a person, object or moment, ask about it before moving on.
4. Ask about what I believed then versus what I understand now. That gap is where memoir finds its meaning.
5. Respect the off-limits topics completely. If I say "pass", move on without comment.
6. Where my answers describe real, identifiable people in a damaging or private light, note it, as those passages may need changed details, consent or legal advice.
7. After 12-15 questions, stop.
Then summarise: the scenes that emerged (each as a one-line scene idea), the strongest sensory details in my own words, the "then versus now" insights, any passages flagged under point 6, and questions to ask other people who were there. Don't write the memoir prose. The words should be mine.
Fill in: MEMORY_TOPIC WHY_IT_MATTERS OFF_LIMITS
Use it before drafting each memoir chapter. Speak your answers aloud and transcribe them if typing slows you down.
Best for: A complete list of claims to verify before publication
Act as a fact-checker for a non-fiction publisher. Read my draft chapter and build a list of every claim I need to verify. Don't verify anything from memory.
My draft chapter: {{DRAFT_CHAPTER}}
Sources I've already used (list or notes): {{MY_SOURCES}}
Produce a table with these columns: Claim (quoted) | Type (statistic, date, quote, attribution, scientific or medical claim, legal claim, historical fact, claim about a living person) | Risk if wrong (low / medium / high) | Source I've given (if any) | What would verify it (the kind of primary source to check).
Then:
1. List quotes attributed to real people, which need the original source, not a quote website.
2. List claims about living people or organisations that could be defamatory if wrong, flagged as high risk.
3. List statistics without a date or source.
4. Note any place where my wording claims more certainty than my source supports ("proves" versus "suggests").
Don't tell me whether claims are true. Your job is to make sure nothing reaches print unchecked. If a claim is health, legal or financial advice, flag it for review by a qualified expert.
Fill in: DRAFT_CHAPTER MY_SOURCES
Use it on every chapter before it goes to an editor. Check the high-risk rows first.
Blurbs, Queries & Author Marketing
The writing that sells the writing. Critique a query letter, turn chapter summaries into a synopsis, prepare a short-story submission, draft back-cover copy in three styles, and write an Amazon description that follows KDP's description rules. Brainstorm keywords for the seven KDP boxes, write your bio, then plan a newsletter, a launch and social posts built from your own best lines rather than hype.
Best for: Turning a query letter into one agents keep reading
Act as a literary agent who reads queries every day. Critique my query letter and help me rebuild it, keeping my voice and my book's facts.
My current query letter: {{QUERY_LETTER}}
Title, genre, age category, word count and comps: {{BOOK_DETAILS}}
The agent I'm sending it to and what their guidelines ask for: {{AGENT_GUIDELINES}}
Critique it against the usual query structure:
1. Hook: does the first line make me want the second? Quote it.
2. Story: protagonist, what they want, what stands in the way, what they must choose and what happens if they fail. Which are missing?
3. Voice: does the pitch sound like the book?
4. Book details: title, genre, word count and comps in one tidy paragraph.
5. Bio: brief and relevant.
6. Length: count the words; most queries run roughly 250-350 words.
7. Red flags: rhetorical questions, the ending revealed or teased coyly, too many named characters, claims about being the next big bestseller.
Then give a rebuild plan: a paragraph-by-paragraph outline of what each part should do, using my book's details. Offer one alternative hook line built from my own wording. Don't rewrite the whole query; I'll draft it from the plan.
Fill in: QUERY_LETTER BOOK_DETAILS AGENT_GUIDELINES
Check each agent's submission guidelines; they vary. Revise, then run the critique again on the new draft.
Example output · Claude (example written by The Prompt Index)
Sample inputs: QUERY_LETTER = "Dear Agent, Have you ever wondered what you would do if you found out someone you loved had lied to you? In THE SALT LEDGER, a historical mystery, that is exactly what happens to Nell Pascoe. Nell is a bookkeeper in a small Cornish village in 1893. Her father died in a shipwreck three years ago, along with his crewmates Will Hosken, Sam Rowe and Peter Jago. One night she finds something strange in a ledger. As she investigates, she discovers that nothing is what it seems, and she must decide what she is willing to sacrifice for the truth. Will she uncover the secret before it's too late? This book will appeal to fans of bestselling historical crime. It is complete at 92,000 words. I have been writing since I was a child and I am a member of a local writing group. Thank you for your time." BOOK_DETAILS = THE SALT LEDGER, historical mystery, adult, 92,000 words; comps: two recent atmospheric historical mysteries set in coastal communities (titles held by the author). AGENT_GUIDELINES = Query plus first 10 pages pasted into the body of the email; no attachments.
Critique
1. Hook: "Have you ever wondered what you would do if you found out someone you loved had lied to you?" This rhetorical question could open any book. Your most specific asset, salt bought for a boat already at the bottom of the sea, doesn't appear at all.
2. Story: protagonist present (Nell, a bookkeeper, 1893). Want: missing; "investigates" isn't a goal. Obstacle: missing; no antagonist or pressure is named. Choice: present but vague ("what she is willing to sacrifice"). Consequence of failure: missing.
3. Voice: the pitch is generic ("nothing is what it seems", "before it's too late"), while the premise is specific and atmospheric. The query should sound like the book.
4. Book details: title, genre and word count are present but scattered; "fans of bestselling historical crime" isn't a comp. Use your two real titles.
5. Bio: "writing since I was a child" uses words without telling the agent anything. Keep the writing group only if it's relevant, or give one line connecting you to the material.
6. Length: about 150 words. You have room to add stakes and voice.
7. Red flags: two rhetorical questions; four named crewmen (three are unnecessary in a query); an implied bestseller comparison.
Rebuild plan
- Paragraph 1 (hook and setup, 60-80 words): Nell, her job, the salt entry and the date that makes it impossible.
- Paragraph 2 (escalation, 80-100 words): what she wants (the truth about her father before the insurer's final payment), who in the village stands to lose, and what it starts to cost her.
- Paragraph 3 (the choice, 30-50 words): the dilemma in concrete terms, ending on the stakes, not a question.
- Paragraph 4 (book details): title, genre, 92,000 words, "will appeal to readers of [comp 1] and [comp 2]".
- Paragraph 5 (bio, one or two lines), then thanks. Paste the first 10 pages below, as the agent asks.
Alternative hook built from your wording: "Three weeks after the Morwenna sank with her father aboard, someone bought twelve bushels of salt for her, and the entry is in his handwriting."
Best for: A complete, plain synopsis that agents ask for with submissions
You are an editor who writes synopses for agent submissions. A synopsis tells the whole story, including the ending, plainly and in order. It is not a blurb.
My chapter summaries or full plot outline: {{PLOT_OUTLINE}}
Main characters and their arcs: {{CHARACTER_ARCS}}
Required length (for example one page, or 500 or 1,000 words): {{LENGTH}}
Write the synopsis with these rules:
- Third person, present tense, in the order events happen.
- Name only the characters the story can't be told without, in capitals on first mention (the common convention).
- Cover the inciting incident, major turning points, the midpoint, the crisis, the climax and the ending, including how the protagonist changes.
- Reveal the ending and any twist. Agents need to see the story works.
- No rhetorical questions, no teasing, no reviews of the book ("a gripping tale").
- Use only events from my outline. If something important seems missing, list it in a note after the synopsis rather than inventing it.
After the synopsis, give the word count and a list of subplots you left out to fit the length, so I can check none are essential.
Fill in: PLOT_OUTLINE CHARACTER_ARCS LENGTH
Edit the draft in your own voice; agents want to see you can write. Check the length each agent asks for.
Best for: Back-cover copy that sells without spoilers
Act as a book-cover copywriter who knows how readers in my genre browse. Write back-cover copy that sells the book without giving away the ending.
My book (premise, protagonist, central conflict, setting): {{BOOK_SUMMARY}}
Genre, subgenre and tone: {{GENRE_TONE}}
What readers love about books like mine (tropes, feelings, hooks): {{READER_APPEAL}}
Write three blurbs of 150-200 words each:
A. Character-led: opens with the protagonist's situation and voice.
B. Hook-led: opens with a striking one-line tagline, then the situation.
C. Trope-led: signals genre pleasures early for readers who browse by trope.
Rules for every version:
- One protagonist (two for romance), no more than three named characters.
- Set up the conflict and the stakes, end on the dilemma, and never reveal the resolution.
- Match the tone of the book: a cosy mystery shouldn't sound like a grimdark thriller.
- Avoid clichés such as "a journey of self-discovery", "nothing will ever be the same" and "in a world where".
- Use only facts from my summary.
After the three blurbs, give a one-sentence tagline for each, and tell me which version you'd test first and why.
Fill in: BOOK_SUMMARY GENRE_TONE READER_APPEAL
Read the blurbs on the back of the bestsellers in your subgenre, then choose and edit one so it sounds like you.
Best for: A formatted Amazon product description that follows KDP's rules
You are a self-publishing marketer who writes Amazon book descriptions that follow KDP's description rules. Write the description for my book's product page.
My back-cover blurb or summary: {{BLURB}}
Genre, subgenre and the reading experience it offers: {{GENRE_EXPERIENCE}}
Series name and number, if any: {{SERIES_INFO}}
Structure:
1. A bold hook of one or two sentences that works on its own, because only the opening lines show before "Read more" on many screens.
2. The blurb, tightened for screen reading: short paragraphs of two or three sentences, ending on the dilemma.
3. One closing sentence, written as natural prose, that tells readers of my subgenre what experience to expect.
4. A series line, if relevant.
KDP's book-description guidelines don't accept reviews, quotes or testimonials, requests for reviews, prices or availability, time-sensitive information, links or contact details, spoilers, emojis, or lists of keywords or tags. Follow all of these. No bestseller or ranking claims either.
Format with tags KDP's description field supports (<b>, <i>, <p>, <br>, <ul>, <li>). The 4,000-character limit counts the HTML tags, so give the total character count including tags at the end.
Then list anything in my blurb you removed to comply, so I know what changed.
Fill in: BLURB GENRE_EXPERIENCE SERIES_INFO
Paste it into the description field on KDP and preview it. KDP's rules change, so re-check its current description guidelines before publishing.
Best for: Keyword phrases and categories to research for a KDP launch
Act as a self-publishing discoverability consultant. Help me brainstorm search phrases for my book's seven KDP keyword boxes and the categories to research.
My book: {{BOOK_SUMMARY}}
Genre, subgenre, tropes and setting: {{GENRE_DETAILS}}
How I think readers search for books like mine: {{READER_SEARCHES}}
Do this:
1. Generate 25 candidate keyword phrases readers might type into Amazon search, grouped by type: subgenre, tropes, setting, character type, mood, and reader occasion (for example beach read). Use phrases of 2-5 words.
2. Remove anything that breaks Amazon's keyword rules as generally understood: other authors' names, other book titles, claims like "bestseller" or "free", time-limited words like "new", and words that misrepresent the content.
3. Pick the seven strongest, combining related phrases so each box covers as much ground as possible, and explain each choice in one line.
4. Suggest 6-8 categories to research in the Kindle store, and tell me what to check (how many books, how recent the top sellers are, whether my book genuinely fits).
Make clear that you can't see Amazon's search volume or rankings. I'll test each phrase in Amazon's search bar and check the current KDP keyword guidelines before using it.
Fill in: BOOK_SUMMARY GENRE_DETAILS READER_SEARCHES
Type each candidate into Amazon search to see what comes up and what autocompletes. Swap keywords if your book doesn't fit the results.
Best for: Submitting short fiction cleanly, with a plain cover letter
Act as a fiction editor at a literary magazine. Help me prepare a short story submission that follows the magazine's guidelines and lets the story speak for itself.
My story (title, word count and the story or a summary): {{STORY}}
The magazine and its submission guidelines, pasted from its website: {{GUIDELINES}}
My publication credits and relevant background (or "none yet"): {{MY_CREDITS}}
Do this:
1. Guideline check: list every requirement in the guidelines (length, formatting, file type, simultaneous submissions, reading period, fees, cover-letter rules) and mark whether my submission meets it, based on what I've given you. Flag anything you can't confirm.
2. Fit: describe in two sentences how my story fits what the guidelines say the magazine wants. If you know nothing reliable about the magazine beyond the guidelines, say so instead of guessing.
3. Cover letter: draft a short, plain cover letter of under 120 words: greeting, title and word count, a one-line description only if the guidelines invite one, my credits or a simple line if I have none, and thanks. No summary of the plot unless asked for.
4. A checklist I can tick before pressing submit.
Don't edit the story itself.
Fill in: STORY GUIDELINES MY_CREDITS
Paste the magazine's guidelines exactly; they vary widely. Keep a log of where each story is under consideration.
Best for: Bios for queries, back covers and press kits from your real facts
You are a publicist who writes author bios that sound human. Write my author bio from the facts I give you, in my tone.
Facts about me (writing credits, relevant jobs, where I live, interests that connect to my books): {{FACTS_ABOUT_ME}}
My books and genre: {{MY_BOOKS}}
The tone I want (warm, wry, serious, mysterious): {{TONE}}
Write three versions in the third person:
1. Short: 35-50 words, for a query letter, a short-story byline or a conference programme.
2. Medium: 80-120 words, for the back of the book or an Amazon author page.
3. Long: 200-250 words, for a website About page or a press kit.
Then write the medium version again in the first person, for my own newsletter or social media.
Rules:
- Use only facts I've given you. Don't add awards, sales figures, degrees or credits.
- Lead with what's most relevant to readers of my genre.
- One personal detail is enough to make it human; don't list hobbies.
- No "passionate about", "avid reader since childhood" or "lives with her cat and too many books" clichés unless I put them in my facts.
Finish with one question about me whose answer would make the bio stronger.
Fill in: FACTS_ABOUT_ME MY_BOOKS TONE
Keep all versions in one document and update them when you publish something new.
Best for: Starting an author newsletter readers stay on
Act as an author-marketing coach who helps novelists build a newsletter readers want to open.
My books and genre: {{MY_BOOKS}}
Material I could offer as a free reader magnet (bonus scene, prequel novella, short story, setting guide): {{MAGNET_OPTIONS}}
How often I can realistically send emails: {{SEND_FREQUENCY}}
Plan:
1. Reader magnet: compare my options for appeal to my genre's readers and how well each leads into book one. Recommend one.
2. Sign-up line: three versions of a one-sentence offer for the back of my book and my website.
3. Welcome sequence: plan 4 emails. For each give the timing, subject line options (two), purpose, and a bullet outline of content. The emails should deliver the magnet, introduce me as a person, share something behind the scenes from my books, and make one clear book recommendation.
4. Ongoing newsletter: a simple repeating format I can keep up at my sending frequency, with 6 content ideas drawn from my books and process.
Keep the tone in the voice of an author, not a sales funnel. Remind me to follow the consent and unsubscribe requirements of my email provider and local law.
Fill in: MY_BOOKS MAGNET_OPTIONS SEND_FREQUENCY
Write the emails yourself from the outlines; readers sign up for your voice.
Best for: A launch plan sized to your actual time and budget
You are a book-launch strategist for independent and small-press authors. Build a realistic launch plan around my resources.
My book, genre and launch date: {{BOOK_AND_DATE}}
What I have (newsletter size, social channels, budget, ARC readers, local connections): {{MY_RESOURCES}}
Hours per week I can spend on marketing: {{HOURS_PER_WEEK}}
Create a week-by-week plan from eight weeks before launch to two weeks after:
1. Each week: 3-5 concrete tasks with an estimate of time for each, matched to my hours. Don't plan more than I have time for.
2. Advance review copies: when to send, how to find readers in my genre, and a short request message. The request must not offer anything in return for a review or ask for a positive one, in line with retailer review rules.
3. Content: 6 post ideas drawn from the book itself (a setting detail, a character question, a line I'm proud of) rather than "buy my book".
4. Launch week: a day-by-day list.
5. A short list of what to measure (newsletter sign-ups, reviews, sales rank checks) and what to skip.
Mark anything that costs money with an estimated range and say it's optional. Don't promise sales results.
Fill in: BOOK_AND_DATE MY_RESOURCES HOURS_PER_WEEK
Adjust the plan to your real hours. A small plan you finish beats a big one you abandon.
Best for: Promoting a book with its own best lines
Act as a book marketer who promotes novels with the author's own words, not generic hype.
Excerpts from my book (favourite lines, a short scene, dialogue): {{EXCERPTS}}
Genre, tropes and the readers I'm trying to reach: {{GENRE_AND_READERS}}
Platforms I use: {{PLATFORMS}}
Do this:
1. Pick 8 lines or moments from my excerpts that work out of context. Quote them exactly; don't polish them.
2. For each, suggest a post format suited to my platforms (quote graphic, short video caption, "tell me you'd read this" post, trope list) and write a short caption of no more than 40 words that frames the quote without spoiling the plot.
3. Write 5 trope or vibe hooks in the style readers use to recommend books (for example "enemies to lovers, but they're both lying"), each built from my book's actual elements.
4. Suggest 3 recurring post ideas I can repeat every week without running out (for example "one line from chapter X").
5. Flag any excerpt that contains a spoiler.
Keep the captions in my voice and genre tone. Don't invent praise, sales figures or comparisons to famous authors.
Fill in: EXCERPTS GENRE_AND_READERS PLATFORMS
Use your own lines; readers respond to the real voice of the book. Check excerpts for spoilers before posting.