tashiscool/scroll-stop-skills/exploded-scene-prompter

Exploded Scene Prompter

Use this skill when the user wants:

• an object or container visually exploding open
• a structured hero prompt package for image and video generation
• a premium HTML prompt page around the concept
• product, food, vehicle, furniture, or service scenes with components clearly art directed

Core Behavior

• Convert rough creative intent into crisp commercial prompt packages.
• Default to photoreal, premium, ad-ready language unless the user requests a different aesthetic.
• If the user asks for a web or HTML view, use assets/prompt-page-template.html as the output scaffold.
• If the object class is ambiguous, ask one concise clarifying question. Otherwise infer reasonable details and keep moving.
• Treat vague prompts as under-specified. Improve them by making the motion, staging, object list, pacing, and continuity explicit.

First-Step Contract

When the user's input is something like:

Hey using this skill, build me a html view for a moving company which is full of furniture and i want the visual to be the back of that van exploding with all of the items

the first step is not to jump straight to HTML.

The first step is to produce the three prompt blocks:

1. PROMPT A - ASSEMBLED SHOT
2. PROMPT B - DECONSTRUCTED / EXPLODED VIEW
3. PROMPT C - VIDEO TRANSITION (START FRAME / END FRAME)

Then, if the user asked for an HTML view, build the HTML page around those prompts.

See references/moving-van-example.md for the canonical pattern.

Output Package

Produce these sections unless the user asks for fewer:

1. PROMPT A - ASSEMBLED SHOT
2. PROMPT B - DECONSTRUCTED / EXPLODED VIEW
3. PROMPT C - VIDEO TRANSITION (START FRAME / END FRAME)
4. ART DIRECTION NOTES
5. HTML VIEW when the user asks for a page, mockup, or prompt page

Only add PROMPT D - DETAIL / ALT ANGLE when the user explicitly asks for extra still coverage.

Prompt Rules

• Start with the subject, camera angle, and background.
• Name visible components explicitly.
• Keep the scene physically believable even when dramatic.
• Prefer white or restrained studio backgrounds unless the user wants an environmental scene.
• Avoid logos and text unless requested.
• Use a high-end product photography tone, not generic “AI art” language.
• For service businesses, show recognizable operational props. Example: moving van, boxes, rugs, lamp, wrapped art, straps.
• When writing the video prompt, always include:
• START FRAME
• END FRAME
• TRANSITION
• STYLE
• The video prompt should explain timing, order of motion, easing, object-level secondary motion, and camera behavior.
• Default camera guidance for transition prompts is locked-off or nearly locked unless the user asks for camera movement.
• Avoid generic verbs like “explode” or “animate” without specifying how, in what order, and with what physical behavior.

Scene Construction Workflow

1. Interpret the subject

• Identify the hero object or container.
• Identify 6-20 components worth surfacing.
• Group them by category: structural parts, payload, styling props, motion debris.
• Infer missing realism details that make the shot better: packing materials, hinges, straps, fasteners, wrappers, cables, labels, soft goods, or hardware.

2. Decide the visual mode

• Assembled: clean intact subject, organized contents visible.
• Exploded: parts suspended in controlled motion, readable spacing, premium ad feel.
• Video transition: a motion plan that transforms the assembled frame into the exploded frame with believable choreography and pacing.

3. Write the prompts

• Use complete, polished paragraphs.
• Mention aspect ratio when the user implies a format like hero banner, landing page, or social asset.
• Add camera, lens, lighting, and quality cues at the end.
• For the video transition, make the transformation directional and staged:
• what moves first
• what follows
• how long the pause and reveal last
• how individual objects rotate, drift, unfold, or separate
• how the final hold behaves

4. Upgrade weak prompts

If the user's original concept is generic, strengthen it before output:

• add a fuller item inventory
• add materials and texture cues
• add motion sequencing
• add secondary motions such as straps unfurling, rugs unrolling, lids opening, or soft items flexing
• add “same lighting setup” and continuity wording between stills and video

5. Build the HTML view when requested

• Read references/prompt-patterns.md for scene formulas.
• Read references/moving-van-example.md when the request is similar to a vehicle or logistics explosion concept.
• Read assets/prompt-page-template.html.
• Replace placeholders with the generated prompt package and scene summary.
• Keep the page elegant and readable, optimized for copy/paste into image tools or creative reviews.

Default Output Style

• Tone: premium commercial art direction
• Format: easy to scan, copy, and refine
• Detail level: rich enough to generate strong images without becoming rambling
• Motion detail: explicit enough that a video model has a real choreography to follow

Example Use Cases

• “Make a laptop exploded into chips, ports, keys, and screen layers.”
• “Create a tropical smoothie burst with fruit, ice, and liquid arcs.”
• “Build an HTML prompt page for a moving company van with furniture exploding out the back.”
• “Give me all three prompts so I can generate the stills first and then the video transition.”

Special Notes

• If the user names a real brand, keep identifiers generic unless they explicitly want brand-accurate styling.
• If motion would become chaotic, bias toward controlled suspension rather than debris chaos.
• When writing furniture or logistics scenes, emphasize order, packing realism, and believable load-out details.
• The third prompt is not a throwaway. Treat it as the bridge between the first two prompts, with clear continuity from still to motion.