seedance-epic-fantasy-scifi

Epic fantasy and sci-fi spectacle · 奇幻科幻大场面

Monsters, dragons, world reveals. Each entity gets its own definition block, the shots are cut by timecode, and scale is bought with reference objects and low angles rather than words like massive.

This Skill carries one prompt structure distilled from 49 human-verified Seedance cases on goodcase.ai. It is a sibling of seedance-prompt-library (all templates in one Skill); install this one when you only want this kind of clip. Do not invent structure from general video-generation knowledge: follow the structure below and ground the draft in one anchor case from references/cases.md.

Use when

• Kaiju attacks, dragon battles, transformation sequences, world reveals — anything whose payload is one physical spectacle at blockbuster scale.
• Chinese: Monster attacking the city, giant dragons battling, transformation sequences, worldbuilding showcase — any payload that is a blockbuster-level physical spectacle.

Workflow

1. Collect inputs. Ask only for what is missing: subject or product, setting, duration, aspect ratio, whether reference images / audio exist, and which language the final prompt should be in.
2. Pick one anchor case. Read references/cases.md, choose the case closest to the request, and name it in the reply. One anchor, never an average of several.
3. Fill the structure block by block. Every block below must be present in the final prompt; an empty block is where prompts go vague.
   1. Opening line: duration, the genre named (cinematic dark fantasy, kaiju action sequence), and whether it is photorealistic or animated
   2. Entity blocks: one paragraph each for the character, the creature, the vehicle and the city, each tagged for what it supplies
   3. Setting and mood: weather, light sources, level of destruction, colour grade direction
   4. Shot breakdown: CUT 1 / CUT 2 with timecodes, or one continuous camera path with named passages
   5. The spectacle beat written on its own, with cause and effect spelled out
   6. Technical tail: grade, fog, grain, lens, render style
   7. Rules paragraph: references are appearance only, keep the face consistent, and state what the final frame holds
4. Apply the guidance and check the pitfalls (next two sections) before returning anything.
5. Return the finished prompt as one copy-pasteable block, the anchor case link, and a three-line checklist of what to verify in the generated video.

Guidance (from the cases)

• Split the entities into their own blocks before any scene text. The kaiju jet case defines Pilot, Seabaycity, Monster and Jet separately, each closed with its scope: Appearance only, Environment only, Vehicle only.
• Buy scale with reference objects and camera angles. The same case instructs sell the size of the monster with low angles and the city for scale, and puts a dramatic low angle in CUT 1.
• Give the clip one spectacle and let the rest be approach. The warrior and white dragon case spends its whole runtime on inserting a key, the dragon emerging, and one beam shattering the celestial orb.
• Write the camera path as something executable. The rain-alley transition case names what it goes through every time: pushes directly toward the center of the ripple, then into the pupil, then through the crystal's internal structure.
• Close with a fixed technical paragraph. The top cases all end on a run like volumetric fog, photorealistic visual effects, Unreal Engine 5 render style with an explicit grade, such as dark grey and golden.

Pitfalls

• Stuffing five spectacles into fifteen seconds. Each one comes out half-finished; give the clip one blow-up and treat every other shot as setup.
• Relying on epic and massive alone. Without a building, a city or a low angle for comparison the monster ends up human-sized.
• Leaving stray characters in the body text. The ruined-bedroom case has a t.co link sitting mid-sentence, and junk like that gets read as picture content — strip it.
• Not saying what the last frame holds. The kaiju jet case ends with Final frame on the monster crashing into the bay, stable and clean; without that line the tail tends to wobble or smear.

Language

Reply in the user's language. The prompt itself can stay in English when that is what the user's Seedance workflow expects; ask if unclear. The Chinese version of this structure follows.

Chinese: Fantasy and sci-fi spectacle

Monsters, giant dragons, worldbuilding showcase. Give each entity its own definition block, cut the shots by timecode, and buy the sense of scale with reference objects and low camera angles — you can't buy it with words like massive.

Structure:

1. Opening line: state the duration and genre (cinematic dark fantasy, kaiju action sequence, etc.), and whether it's live-action or animation
2. Entity definition blocks: one paragraph each for characters, monsters, vehicles, and the city, with each paragraph clearly stating what it alone provides
3. Scene and atmosphere: weather, light sources, degree of destruction, color grading direction
4. Shot breakdown: CUT 1 / CUT 2 with timecodes, or a single continuous shot path that names what it passes through each time
5. Write the spectacle beat separately, and make the cause and effect clear
6. Technical closing paragraph: color grading, fog, grain, lens, render style
7. Rules paragraph: reference images are for appearance only, faces must stay consistent, and what the final frame lands on

Key points:

• First break the entities into separate blocks, then write the scene. In the fighter jet monster one, write a separate paragraph each for Pilot, Seabaycity, Monster, and Jet, and state the purpose at the end of each paragraph:Appearance only、Environment only、Vehicle only。
• The sense of scale is bought with reference objects and camera angles. The same one explicitly requires sell the size of the monster with low angles and the city for scale, and CUT 1 is a dramatic low angle.
• Give the whole thing only one spectacle; all the other shots are just positioning. In the female warrior white dragon one, the entire film is just inserting the key, the dragon appearing, and a single beam of light shattering a celestial body.
• Write the shot path as executable actions. In the rainy alley transition one, name what it passes through at every scene change:pushes directly toward the center of the ripple, then pass into the pupil, then through the crystal's internal structure.
• Always close with a paragraph of technical parameters. The most popular ones all end on a string like volumetric fog, photorealistic visual effects, Unreal Engine 5 render style , and specify the color grading direction, such as dark gray plus gold.

Common pitfalls:

• Cramming five spectacles into fifteen seconds. Each one ends up half-finished; give a single video only one explosive moment, and treat the other shots as buildup.
• Relying only on words like epic and massive. With no buildings, no city, and no low camera angle for reference, the monster comes out the same height as a person.
• Leaving dirty characters in the body text. In the ruined city black cat one, a t.co link is wedged in the middle of a sentence; that kind of thing gets treated as image content and must be cleaned out.
• Not stating where the final frame lands. The fighter jet monster one wrote Final frame on the monster crashing into the bay, stable and clean; without that closing line, it easily shakes or blurs.

Copy-ready lead-in / 可复制引导语

For users who would rather paste into a chat model than run this Skill, hand them this line above the structure:

I want an epic fantasy spectacle clip. [The lead is a young man with a long blade on his back, standing in a half-buried desert city.] [The spectacle: a sand worm bursts up through the street from below.] Using the prompt template below, rewrite it into one ready-to-use Seedance video prompt for me:

I want to make a fantasy spectacle video, [the protagonist is a boy carrying a long blade, standing in a city half buried by sand], [the spectacle is a sandworm bursting up through an entire street from underground]. Please use the prompt template below to help me rewrite it into a Seedance video prompt I can use directly:

Notes

• references/cases.md is generated from data/ in LearnPrompt/awesome-seedance and refreshed daily; do not hand-edit an installed copy.
• Prompts in the reference file belong to their creators (linked on every card). Transfer the structure; do not present a close copy as original work.
• For the full library across every model, install the goodcase Skill from LearnPrompt/goodcase-lite.