Claude Prompt Library

Claude for Authors: The Writing Partner That Actually Reads Your Whole Manuscript

20 copy-paste prompts

20 Claude prompts for plot structure, character arcs, scene-level craft, query letters, book marketing, and the long-form coherence authors need.

In short: This page contains 20 copy-paste ready prompts, organized into 4 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

Louis Corneloup
By Louis Corneloup · Founder, Techpresso
Last updated ·Hand-curated & tested by the AI Academy team

Story Development

5 prompts

Three-Act Outline

1/20

✨ What it does

Outlines novels in three-act structure with scene-level beats and genre-specific pitfalls.

<task>Outline novel in three-act structure</task> <genre>[describe]</genre> <premise>[1-2 sentences]</premise> <protagonist>[goal, flaw, stake]</protagonist> <output> Act 1: inciting incident, first plot point, 3 key scenes Act 2a: rising action, midpoint reversal Act 2b: crisis, all-is-lost Act 3: climax, resolution, 3 key scenes Theme + genre pitfalls.

💡

Pro tip: Claude's long-context + structured reasoning handle novel outlines with coherence. For maintaining plot integrity across multi-part outlines, Claude outperforms ChatGPT.

Save the Cat Beat Sheet

2/20

✨ What it does

Builds Save the Cat beat sheets with 15 beats, page targets, and emotional function.

<task>Save the Cat beat sheet</task> <genre>[describe]</genre> <logline>[1 sentence]</logline> <output> 15 beats (Opening Image through Final Image). Per beat: page target, what happens, emotional function.

💡

Pro tip: Save the Cat maps to how humans process story. Use as diagnostic. If your midpoint feels flat, beat sheet reveals what's missing page 50-55.

Character Arc Design

3/20

✨ What it does

Designs character arcs with internal milestones, belief shifts, and mirror characters.

<task>Design protagonist arc</task> <character>[describe]</character> <start_state>[flaw, belief, pattern]</start_state> <end_state>[who they become]</end_state> <output> 3-5 internal milestones + external triggers + "lie they believe" vs "truth to learn" + relapse moment + final test + supporting character mirrors/challenges.

💡

Pro tip: Arcs without internal transformation = sequence of events. Readers want to watch someone change. Same person at end = plot puppet, not protagonist.

World-Building Blueprint

4/20

✨ What it does

Builds worlds with central rule, implications, and strategic info reveal.

<task>Build world for [genre]</task> <premise>[describe]</premise> <output> 1. Central rule (one thing different from ours) 2. 5 implications of that rule 3. Geography + setting 4. Social structure + power dynamics 5. Economy 6. Conflict engine 7. Chapter 1 vs later reveals 8. Iceberg principle </output>

💡

Pro tip: Great world-building: one fundamental rule + its consequences. "Magic exists but costs memory" gives you 100 plot hooks. Iceberg: know 10×, reveal 1×.

Subplot Weaver

5/20

✨ What it does

Integrates subplots with main plot using theme, character, and pacing integration.

<task>Weave subplots into main plot</task> <main_plot>[describe]</main_plot> <characters>[list]</characters> <themes>[describe]</themes> <output> 3 subplots serving main story: character driver, conflict created, theme connection, climax moment, pacing, interaction between subplots, cut priority.

💡

Pro tip: Subplots must serve main story. Not intersecting main plot meaningfully = side quest, cut it. Interwoven subplots compound emotional weight; parallel distract.

Craft & Revision

5 prompts

Scene Starter

6/20

✨ What it does

Generates 3 scene openers in different styles with emotional alignment.

<task>Write scene starter</task> <scene_context>[describe]</scene_context> <pov>[character]</pov> <goal>[plot movement]</goal> <emotion>[describe]</emotion> <output> 3 opening paragraph options (action / dialogue / setting), for each: emotional beat fit, what to avoid, POV character state cue.

💡

Pro tip: Most scenes start too early. Cut hero walking in, looking around, ordering coffee. Start where something changes. Boring openings usually need to start one page later.

Dialogue Polisher

7/20

✨ What it does

Polishes dialogue for subtext, distinct voices, action beats, and strategic silence.

<task>Polish dialogue</task> <dialogue>[paste]</dialogue> <scene_context>[describe]</scene_context> <characters>[describe]</characters> <problems>[on-the-nose / stilted / info-dumpy]</problems> <output> Revised dialogue with: subtext, distinct voices, action beats (not just "he said"), strategic silence, emotional subtext.

💡

Pro tip: Real people don't say what they mean. Fictional dialogue that says what character means = flat. Characters speak AROUND their truth, not at it. Reader hears what's not said.

Show-Not-Tell Fixer

8/20

✨ What it does

Converts telling to showing with physical reaction, characterization through behavior, and sensory mood.

<task>Fix telling in passage</task> <passage>[paste]</passage> <output> Rewrite showing: internal emotion through physical reaction, character judgment through action/word choice, backstory through present behavior (no flashback dumps), setting through sensory detail revealing mood. Flag what to never telegraph directly.

💡

Pro tip: "She was angry" tells. "She flipped the kitchen chair before slamming the door" shows. All showing exhausts — balance. Show what matters emotionally; tell connective tissue.

Sensory Layering

9/20

✨ What it does

Layers specific sensory detail with mood-reinforcing selection and restraint.

<task>Layer sensory detail into scene</task> <scene>[paste]</scene> <mood>[describe]</mood> <output> Add: specific sights, sounds (ambient + specific), smells (one strong, one subtle), physical sensations on POV character, taste if relevant. Select details that carry weight.

💡

Pro tip: All senses at once = travel brochure. 2-3 senses per scene. Smell is underused superpower — memories/emotions connect through scent faster than sight.

Pacing Diagnosis

10/20

✨ What it does

Diagnoses pacing with tension meter, cut/expand list, and chapter hooks.

<task>Diagnose pacing</task> <section>[paste or summarize]</section> <word_count>[number]</word_count> <output> 1. Where pacing drags 2. Where it rushes 3. Scene tension meter (1-10 each) 4. What to cut / expand 5. Chapter-ending hooks 6. Genre expectation comparison </output>

💡

Pro tip: Pacing isn't speed — feeling of forward motion. Slow scene feels fast if something's at stake; fast scene feels slow if nothing matters. Tension > action.

Query & Publishing

5 prompts

Query Letter Draft

11/20

✨ What it does

Drafts query letters with hook, plot, closing, bio, and housekeeping in industry format.

<task>Draft query letter</task> <title>[title]</title> <word_count>[number]</word_count> <genre_age>[describe]</genre_age> <comps>[2-3]</comps> <bio>[describe]</bio> <logline>[1 sentence]</logline> <output> Hook paragraph (protagonist + incident + stakes) → plot paragraph (conflict, choice, stakes) → closing (emotional core) → bio → housekeeping. Under 350 words.

💡

Pro tip: Agents decide on queries in 60 seconds. First paragraph must hook with protagonist + trouble + stakes. Save world-building for manuscript. Generic "In a world where..." dead on arrival.

Synopsis Writer

12/20

✨ What it does

Writes synopses with full plot, emotional arc, turning points, and ending reveal.

<task>1-2 page synopsis</task> <plot>[full plot + ending]</plot> <characters>[list]</characters> <output> 500-750 words covering full plot with ending. Emphasize character motivation + emotional arc. Major turning points. Present tense, third person. Cut subplots not serving main thread.

💡

Pro tip: Synopses: most hated part of publishing, most essential. Agents use them for craft assessment. Show emotional arc, not just events. Most read like plot summary; great ones read like emotional journey.

Book Blurb / Back Cover

13/20

✨ What it does

Writes 150-200 word back-cover blurbs with emotional hooks and tone matching.

<task>Back-cover blurb</task> <title>[title]</title> <genre>[describe]</genre> <protagonist>[describe]</protagonist> <conflict>[describe]</conflict> <tone>[describe]</tone> <output> 150-200 words: intriguing opener → protagonist + situation → inciting incident + conflict → stakes → spoiler-free tease. Match book's actual tone.

💡

Pro tip: Great blurbs promise what book delivers. Blurb promising romantic suspense + book is slow literary = readers feel cheated. Match blurb's tone to book's tone.

Comp Title Research

14/20

✨ What it does

Finds realistic comp titles with recent publication, similar tone, and differentiation reasoning.

<task>Find comparable titles</task> <my_book>[premise + tone + audience]</my_book> <output> Criteria: 3-5 years recent, same genre+age, similar tone/aesthetic, moderate success (not mega-bestseller). 2-3 ideal comps + 1-sentence justification + how my book is similar yet distinct.

💡

Pro tip: Comp failures: too big ("Harry Potter meets Gone Girl") or too old (2010 books). Find moderately successful 2022-2026 books sharing tone. Comps signal you understand market.

Author Bio for Queries

15/20

✨ What it does

Writes 3 author bio variations for queries — credential, platform, personality emphasis.

<task>3 author bios for queries</task> <credentials>[list]</credentials> <platform>[describe]</platform> <output> Each 2-3 sentences: lead with most impressive credential, avoid irrelevant trivia, mention platform, memorable detail. No "always loved writing." 3 versions: credentials, platform, personality emphasis.

💡

Pro tip: Debut authors no credits = emphasize relevant life experience. What makes YOU the person to write this book? Unique experience beats generic passion.

These prompts give you the what. Tutorials give you the why.

Learn when to use extended thinking, how to build Claude Projects, and workflows that compound. 300+ tutorials and growing.

Try AI Academy Free

Marketing

5 prompts

Book Launch Plan

16/20

✨ What it does

Plans book launches with 90-day pre-launch, launch week, post-launch, and measurement.

<task>Plan book launch</task> <genre>[describe]</genre> <audience>[describe]</audience> <launch_date>[date]</launch_date> <platform>[current]</platform> <output> 1. 90-day pre-launch (ARCs, reviews, cover reveal, preorder) 2. Launch week 3. 30-day post-launch 4. Playlist pitching 5. Press outreach 6. Realistic sales expectations 7. Success metrics </output>

💡

Pro tip: Launches aren't events — 90-day sustained campaigns. Most front-load launch week and vanish. Winners build momentum 30-60 days AFTER release through reviews, content, collabs.

Author Newsletter Strategy

17/20

✨ What it does

Builds author newsletter strategy with pillars, lead magnet, and cross-promos.

<task>Author newsletter strategy</task> <list>[size]</list> <genre>[describe]</genre> <goals>[sell books / loyal readers / paid tier]</goals> <output> 1. Content pillars (author life, craft, recs, exclusive, preview chapters) 2. Frequency 3. Lead magnet 4. Paid tier options 5. Launch leverage 6. Cross-promos with same-genre authors </output>

💡

Pro tip: Newsletters with only "buy my book" die fast. Content serving readers first builds loyalty. When you launch, engaged list buys because they love the relationship.

Social Media for Authors

18/20

✨ What it does

Builds author social strategy with platform prioritization and burnout-safe cadence.

<task>Author social strategy</task> <genre>[describe]</genre> <time>[hours/week]</time> <output> 1. Best platform for genre (TikTok/BookTok, Bookstagram, Twitter, YouTube) 2. Content pillars (process, bookish, personality, marketing) 3. Posting frequency 4. BookTok/Bookstagram tactics 5. Avoiding burnout 6. KPIs correlating to sales </output>

💡

Pro tip: Authors burn out trying to be everywhere. Pick ONE platform matching energy + genre. TikTok for broad fiction, Substack for literary, Twitter for SFF. Go deep on one.

Book Review Outreach

19/20

✨ What it does

Plans book review outreach with reviewer sourcing, templates, and ARC kits.

<task>Plan book review outreach</task> <genre>[describe]</genre> <book>[brief]</book> <output> 1. Reviewer sourcing (Goodreads, BookTok, bookstagram, trade) 2. Outreach template 3. ARC offer + logistics 4. Review kit (cover, blurb, samples, bio) 5. Follow-up etiquette 6. Gift ethics </output>

💡

Pro tip: Reviewers get 100+ requests. Winning pitches come from genuine fans of their content. Follow 10-20 reviewers for 3+ months before asking. Authentic relationships > cold email.

Series Strategy

20/20

✨ What it does

Plans series strategy with arc structure, continuity, release cadence, and retention.

<task>Plan series strategy</task> <book1_premise>[describe]</book1_premise> <total_books>[number]</total_books> <output> 1. Overarching plot vs book-specific 2. Character arcs (span vs per-book) 3. World-expansion cadence 4. Ending book 1 to hook book 2 5. Continuity management (Bible / wiki) 6. Release schedule (rapid vs annual) 7. Retention between books </output>

💡

Pro tip: Series with strong book 1 endings sell book 2. Vague "stay tuned" loses readers. End book 1: main conflict resolved + new thread opened. Satisfying + hungry = book 2 buyers.

Free tool

Prompt Optimizer

Turn a rough idea into a structured, professional AI prompt.

Try it free →

Frequently Asked Questions

No — fiction that connects emotionally requires lived experience and voice. But Claude will replace generic content. Authors using Claude as craft assistant (outlining, editing, research) outproduce those who don't. AI + human voice = leverage; AI alone = forgettable.
Depends on use. Brainstorming, outlining, editing = fine (like co-writing). AI-generated scenes claimed as yours = gray area. Best practice: use Claude to accelerate YOUR creativity, not replace it.
Yes — finishing is structure + momentum, not raw writing. Claude diagnoses why you're stuck, works through plot holes, writes through blocked scenes. Many authors stalled for years finish drafts with Claude as thinking partner.
Claude's long-form coherence + voice matching are author-specific strengths. For maintaining manuscript voice across 80,000 words, Claude wins. ChatGPT faster for short ideation.
Depends on goals + book. Traditional: prestige, bookstore placement, advances, slow (18-24 months), less control. Self-pub: speed, control, higher royalty %, full marketing burden. Research your genre's norms.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

7-day free trial. Cancel anytime.