A playable dive into an ancient computer
A one-minute capture of a procedural historical game turns underwater exploration into an interactive artifact reveal.
Starter prompt / by Refs
The creator shared no short prompt. This is ours for making a film like it with Claude.
Build a small playable game in one HTML file with three.js, inspired by {HISTORICAL_OBJECT}: dive underwater to find it, with schools of fish, swaying seagrass and floating particles, then inspect the object as a geared model you can turn. Generate every visual and sound in code with no external assets, and keep it smooth in the browser.Field notes / why it works
The opening 2.65 seconds show a complete playable premise, controls and a boat before the camera plunges underwater. A long continuous game capture lets the viewer discover and handle clues (0:14–0:30), then translates the object into a luminous diagram at 0:31–0:34 and a readable payoff over the returning boat at 0:44–0:55.
Plate 01 / The format in one breath
A 60-second recording starts on the title screen of a playable historical world, invites a single action, follows the player into a discovery, and ends by explaining why the discovered object matters. The promise is visible at 0:00 in the title, 1901 setting and “BEGIN THE DIVE” control; the answer is held until about 0:42–0:47.
Why it works
- The first frame looks playable: a small sailboat, period setting, one clear button and a control legend (0:00–0:02.65).
- Exploration replaces explanatory cuts. The diver moves from surface to seabed, touches fragments and handles a device across the 0:02.65–0:35.6 continuous game capture.
- It translates the artifact into a consequence: an orbital diagram at 0:31–0:34, then a large discovery line over the returning boat at 0:44–0:55.
Plate 02 / Format card
| Platform · aspect · length | X · 900 × 720, 5:4 · 60.0 s · 60 fps |
| Pace | 2 detected shots; 1 hard cut = 1/min; 8 changes within the long shot = 9 picture changes/min. The second detected shot lasts 57.32 s. Game camera and scene motion make the viewing rhythm much faster than that cut count suggests. |
| Script | Almost no spoken copy: one low-confidence three-word phrase at 0:26.1. Title cards, in-game dialogue and controls carry the story. No reliable narration word budget. |
| Voice | Brief, low-register in-game-sounding utterance, if the transcription is accurate; no sustained narrator. |
| On screen | About 3 s title/menu, 33 s third-person underwater interaction, 7 s dark transition/reveal, 17 s boat/end card. All appear to be captured from the same interactive world. |
| Captions | No social-video karaoke captions. Small in-world dialogue/subtitles, serif opening and closing titles, and tiny game HUD. |
| Sound | Sound is present per the post. Measured −18.7 LUFS integrated, −1.8 dBFS true peak; a music-like pulse near 150 BPM from 0:29.6 onward is a rough detector result, not confirmed instrumentation. |
| Structure | 5 beats: invite → enter world → search/inspect → mechanism reveal → outcome/end screen. Payoff 0:31–0:47; replay invitation from about 0:53. |
Pro / 13 plates + recipe
The first plates are open. The full dissection is Pro.
Get the timed shots, type and sound specs, and the runnable recipe.
Open the full blueprint- Plate 03 / The hook, frame by frame (first 3 seconds)Pro
- Plate 04 / Structure (beat sheet)Pro
- Plate 05 / Shot grammarPro
- Plate 06 / Visual style guidePro
- Plate 07 / Voice and deliveryPro
- Plate 08 / Sound designPro
- Plate 09 / PackagingPro
- Plate 10 / Script templatePro
- Plate 11 / Production planPro
- Plate 12 / PromptsPro
- Plate 13 / Edit recipePro
- Plate 14 / QA checklistPro
- Plate 15 / Make it yoursPro
- Recipe / runnable promptPro