Verified against ChatGPT · 2026-07-30
Structure a storytime video with a real narrative arc, not just chronological retelling
Rebuilds a personal-story video around what the storyteller understood by the end versus the start, with deliberate tension placement — instead of a flat, chronological "and then this happened" retelling that has no reason to hold attention.
The prompt
Ready to copy — highlighted parts are example details you can swap.
You are structuring a "storytime" narrative YouTube script — a first-person account of something that actually happened, told for an audience, not a court deposition. A story that just recounts events in the order they occurred, with no shaping, has no reason to hold attention past the first factual surprise; you are building an actual narrative arc on top of the true events, without inventing anything that didn't happen. WHAT ACTUALLY HAPPENED Moved in with a roommate found online, seemed normal for two months, then discovered they'd been forging the storyteller's signature on a shared lease addendum to add a third undisclosed occupant WHAT YOU UNDERSTOOD DIFFERENTLY BY THE END Believed being a 'chill, trusting' roommate was a virtue; ended up realizing that specific trust was what let two months of forged paperwork go unnoticed TONE Mostly matter-of-fact and a little self-deprecating about being naive, not outraged or played for pure shock TARGET LENGTH 13-16 minutes SENSITIVE DETAILS TO HANDLE CAREFULLY Don't name the roommate's actual first name or city — refer to them only as 'the roommate' NARRATIVE RULES Identify the actual dramatic question the story is answering — not "what happened" but the specific thing the audience is meant to wonder throughout: will this person get away with it, was the storyteller right to trust someone, how did an ordinary decision turn into a disaster — and structure every scene to either build or complicate that question, not just to report the next chronological event. Do not reveal the ending's key turn earlier than the story earns it — if the true events include an early detail that gives away the ending's twist, either withhold restating its significance until later, or use it as deliberate foreshadowing the audience only recognizes in hindsight, rather than accidentally spoiling your own story's turn three minutes before it should land. Include the storyteller's real, in-the-moment belief or assumption at each stage, not just what's now known in hindsight — the tension of a good storytime often comes specifically from the gap between what the narrator believed then and what the audience, hearing it told after the fact, can already sense is about to go wrong; collapsing that gap by narrating everything with hindsight-informed certainty removes exactly the suspense that gap creates. Never invent a detail, quote, or event that didn't happen to heighten drama — if a real detail is genuinely unclear or forgotten, say so honestly in the script rather than fabricating a specific one, since a storytime audience's trust is the entire product and a single caught fabrication collapses it retroactively across the whole channel. Handle any sensitive detail specified with the exact level of specificity or vagueness requested — do not add identifying detail beyond what's given, and do not sensationalize a genuinely difficult moment past what the storyteller is actually comfortable sharing. OUTPUT FORMAT 1. One line stating the dramatic question the whole story is built to answer. 2. The full script broken into scenes with approximate timestamps, written in first person, spoken lines. 3. A note on where the in-the-moment belief versus the hindsight reveal are deliberately kept separate, and where. 4. Confirmation that no invented detail was added beyond what was provided as true.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Naming the actual dramatic question before writing a single scene is what separates a shaped narrative from a chronological report of the same true events: "what happened" is a sequence of facts with no inherent reason to keep watching once the facts are known, while a specific question like "was trusting this person a mistake, and when did it become one" gives every scene a job — build the question or complicate it — that a purely chronological retelling has no equivalent mechanism for, even when both versions describe the identical real events. Deliberately preserving the gap between the storyteller's in-the-moment belief and their current hindsight knowledge targets the single most common flattening effect that happens when a true story gets narrated by someone who already knows how it ends: without a specific instruction to hold that gap open, the retelling naturally collapses into hindsight-informed certainty at every stage — "of course I should have seen the forged signature coming" — which removes exactly the tension a listener gets from knowing more than the past version of the narrator did at each point, the mechanism most true-crime and storytime narration actually runs on. The rule against fabricating detail to heighten drama is not merely an ethics instruction bolted onto a creative task — it's protecting the actual asset a storytime channel depends on, since the format's entire appeal rests on the audience's belief that this really happened, and a single caught invented detail doesn't just damage that one video, it retroactively makes every other story from the same channel suspect once an audience has reason to wonder what else was embellished. The explicit instruction to say "this part is unclear" rather than confidently fabricating a forgotten specific matters for the identical reason from a different angle: a script with a small, honestly-flagged gap reads as more credible, not less, than one with suspiciously perfect recall of a years-old conversation's exact wording, because real memory doesn't work that way and audiences who consume a lot of this format have learned to distrust the too-perfect version.
Verified against
ChatGPT GPT-5.1 · 2026-07-30
Claude Sonnet 4.6 · 2026-08-01
Changelog
- 2026-07-30 — Initial publish, verified against ChatGPT (GPT-5.1) and Claude (Sonnet 4.6).
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
