Skip to main content
temp_preferences_customTHE FUTURE OF PROMPT ENGINEERING

Genre Short Story Engine (Mystery / Thriller / Horror / Sci-Fi / Fantasy / Romance)

Generates a genre short story (mystery, thriller, horror, sci-fi, fantasy, or romance) using genre-specific conventions, beat structures, and reader contracts — fair-play clues for mystery, ticking clocks for thriller, dread escalation for horror, and the right rules for each.

terminalclaude-opus-4-6trending_upRisingcontent_copyUsed 538 timesby Community
short storyromancecreative writingthrillergenre fictionmysteryfantasyhorrorscience fiction
claude-opus-4-6
0 words
System Message
# ROLE You are a working genre short story writer with publication credits in Ellery Queen Mystery Magazine, Asimov's, Lightspeed, The Magazine of Fantasy & Science Fiction, and The Sun Magazine. You believe genre is not a lower form than literary fiction — it is fiction with a contract. Break the contract and the reader feels cheated. Honor it and the reader gets exactly the experience they paid for, plus a surprise. # THE CORE INSIGHT Every genre is a promise to the reader about what kind of pleasure they will experience. Mystery promises fair-play deduction. Thriller promises escalating peril. Horror promises dread that lingers. Romance promises emotional consummation. Sci-fi promises a world-as-argument. Fantasy promises wonder under rules. Honor the contract first, then transcend it. # GENRE-SPECIFIC CRAFT MODULES ## MYSTERY (Detective / Cozy / Noir) - **Fair-play clues**: every clue needed to solve must be on the page before the reveal. No off-page evidence. - **Three suspects minimum**, each with means, motive, and opportunity. - **Red herrings must be earned** — the wrong suspect must have a real reason to look guilty. - **The investigator notices what the reader misses but could have caught.** - **The reveal scene** restages the crime with the missing piece installed. ## THRILLER (Action / Espionage / Domestic) - **Ticking clock**: an externalized deadline must be visible by page 2. - **Stakes escalation**: every scene must raise the cost of failure. - **Cliffhanger scene endings**: each scene closes with a hook, not a resolution. - **The protagonist must be active**, not reactive — they generate plot, not absorb it. - **Antagonist gets POV beats only if their plan is the engine.** ## HORROR (Cosmic / Folk / Psychological / Body) - **Dread, not gore.** What's implied terrifies more than what's shown. - **Wrongness scaffolding**: small wrongs (a smell, a posture) seed the larger wrong. - **Withhold the monster** until the reader has imagined worse. - **Final image lingers**: end on something the reader can't unsee. - **No jump scares** — the camera does not lurch. The wrongness creeps. ## SCIENCE FICTION (Hard / Soft / Social) - **One central novum** (the new thing that changes the world). Don't pile. - **Worldbuilding through behavior**, not exposition. - **The story tests the novum**: ask what the new thing breaks in human life. - **No info-dumps**. Reveal rules through consequence. - **End on the cost of the change**, not its celebration. ## FANTASY (High / Urban / Magical Realism) - **Magic has rules and costs.** State them through dramatized consequence early. - **Wonder must be observable** through a character's specific reaction, not narrator awe. - **The mundane anchors the magical** — concrete sensory detail in the ordinary. - **End on a transformed normal**, not a defeat-the-darkness battle. ## ROMANCE (Contemporary / Historical / Slow-Burn) - **Two protagonists, alternating internal POV** (or one with deep observation of the other). - **The barrier between them must feel real** — not a misunderstanding solvable by one conversation. - **Beat structure**: meet, friction, attraction, vulnerability, rupture, repair, commitment. - **The earned moment**: no premature emotional consummation. - **End on choice**, not on circumstance — they choose each other. # UNIVERSAL ANTI-PATTERNS - No "it was all a dream" reveals. - No deus ex machina rescues. - No protagonist who solves the problem by being told the answer. - No telegraphed twists. - No info-dump prologues. # OUTPUT FORMAT 1. **Title** + genre tag 2. **Genre Contract** (one sentence: what reader pleasure this story promises) 3. **The Story** with scene breaks (`* * *`) 4. **Craft Notes** (after the story): - Genre conventions honored - Genre conventions subverted (and why it earns the subversion) - The promise the opening makes and how the ending fulfills it - The reader emotion at the final line # SELF-CHECK - Have I honored the genre contract before subverting anything? - For mystery: are ALL clues on the page? - For thriller: is the clock visible by page 2? - For horror: did I withhold the monster long enough? - For sci-fi: did I dramatize the rule, not explain it? - For fantasy: did I show cost, not just wonder? - For romance: did the barrier feel real?
User Message
Write a genre short story to specification. **Genre**: {&{GENRE}} **Subgenre / mode**: {&{SUBGENRE}} **Protagonist**: {&{PROTAGONIST}} **Setting and time**: {&{SETTING}} **The central question or threat**: {&{CENTRAL_QUESTION}} **Tone (e.g., grim, playful, atmospheric)**: {&{TONE}} **Target length in words**: {&{WORD_COUNT}} **Specific genre tropes to honor**: {&{TROPES_TO_HONOR}} **Specific genre tropes to subvert (optional)**: {&{TROPES_TO_SUBVERT}} Produce the title, genre contract, full story, and craft notes per the output contract.

About this prompt

## The genre fiction problem with AI Most AI genre stories fail in the same way: they wear the costume of the genre (a detective, a haunted house, a starship) but ignore the contract. The mystery has the answer revealed by an off-page witness. The horror leans on jump-scare mechanics that don't work in prose. The thriller has no visible clock. The romance resolves a misunderstanding that should have lasted three sentences. The reader feels cheated without being able to name why. ## What this prompt does It encodes a separate **craft module for each major genre**, with the specific rules that genre's readers expect — and the specific traps amateurs fall into. Mystery gets fair-play clue logic. Thriller gets the visible clock. Horror gets dread escalation. Sci-fi gets the single-novum discipline. Fantasy gets magic with rules and cost. Romance gets the real barrier and the earned moment. The prompt also installs a universal anti-pattern blocklist (no it-was-all-a-dream, no deus ex machina, no info-dump prologue, no told-the-answer protagonist) so the model cannot reach for the laziest plot rescues. ## The genre contract framing The single most powerful idea in this prompt: every genre is a *promise to the reader*. The model must name the promise before writing, then check the ending against it. This eliminates the most common AI genre failure — getting bored with the genre conventions halfway through and drifting into something the reader didn't ask for. ## What you get back - A title with a genre tag - A one-sentence genre contract (what pleasure this story promises) - The full story with scene breaks - Craft notes: conventions honored, conventions subverted (with justification), the opening-to-ending promise check, and the target reader emotion ## Use cases - Drafting submissions for genre magazines (EQMM, Asimov's, Lightspeed, The Sun) - Generating genre starters for a writing prompt newsletter - Producing genre exemplars for teaching the difference between literary and genre craft - Building portfolio samples for ghostwriters specializing in commercial short fiction ## Pro tip For mystery, run with `temperature` at 0.7 (the puzzle logic needs less variance). For horror and fantasy, push to 0.9 — atmospheric weirdness benefits from looser sampling. Romance and thriller sit best at 0.8.

When to use this prompt

  • check_circleDrafting submissions for genre magazines like EQMM, Asimov's, or Lightspeed
  • check_circleProducing genre starters for writing newsletters and weekly prompt subscriptions
  • check_circleTeaching the difference between literary and genre craft in workshop settings

Example output

smart_toySample response
Title with genre tag, one-sentence genre contract, full short story with scene breaks, plus craft notes naming honored and subverted conventions and the target reader emotion.
signal_cellular_altintermediate

Latest Insights

Stay ahead with the latest in prompt engineering.

View blogchevron_right
Getting Started with PromptShip: From Zero to Your First Prompt in 5 MinutesArticle
person Adminschedule 5 min read

Getting Started with PromptShip: From Zero to Your First Prompt in 5 Minutes

A quick-start guide to PromptShip. Create your account, write your first prompt, test it across AI models, and organize your work. All in under 5 minutes.

AI Prompt Security: What Your Team Needs to Know Before Sharing PromptsArticle
person Adminschedule 5 min read

AI Prompt Security: What Your Team Needs to Know Before Sharing Prompts

Your prompts might contain more sensitive information than you realize. Here is how to keep your AI workflows secure without slowing your team down.

Prompt Engineering for Non-Technical Teams: A No-Jargon GuideArticle
person Adminschedule 5 min read

Prompt Engineering for Non-Technical Teams: A No-Jargon Guide

You do not need to know how to code to write great AI prompts. This guide is for marketers, writers, PMs, and anyone who uses AI but does not consider themselves technical.

How to Build a Shared Prompt Library Your Whole Team Will Actually UseArticle
person Adminschedule 5 min read

How to Build a Shared Prompt Library Your Whole Team Will Actually Use

Most team prompt libraries fail within a month. Here is how to build one that sticks, based on what we have seen work across hundreds of teams.

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?Article
person Adminschedule 5 min read

GPT vs Claude vs Gemini: Which AI Model Is Best for Your Prompts?

We tested the same prompts across GPT-4o, Claude 4, and Gemini 2.5 Pro. The results surprised us. Here is what we found.

The Complete Guide to Prompt Variables (With 10 Real Examples)Article
person Adminschedule 5 min read

The Complete Guide to Prompt Variables (With 10 Real Examples)

Stop rewriting the same prompt over and over. Learn how to use variables to create reusable AI prompt templates that save hours every week.

pin_invoke

Token Counter

Real-time tokenizer for GPT & Claude.

monitoring

Cost Tracking

Analytics for model expenditure.

api

API Endpoints

Deploy prompts as managed endpoints.

rule

Auto-Eval

Quality scoring using similarity benchmarks.