Get the skill

●  Daily on TikTok / Reels / Shorts

The Seedance 2.0 skill used in  @crowdprompt .

We use it every day to make our videos. Now it's yours.

Drop it in ChatGPT, Gemini, Claude, or any other LLM. Describe your scene. Run the output through Seedance. You get a video.

Free. No spam. Just the file in your inbox.

a video we made

"POV world war one pilot dodging missiles"

How to use it

Four steps. No setup. No SDK.

  1. new chat
    MD crowd-prompt-seedance-2.md 10.4 KB · MARKDOWN
    + Ask anything… ↑
    01

    Paste the skill into your chatbot

    ChatGPT, Gemini, Claude, or any other LLM that takes a system prompt. The whole file. One time.

  2. new chat
    MD crowd-prompt-seedance-2.md

    a knight kneeling in a snowy field at dusk, fog rolling in, cinematic

    Ask anything… ↑
    02

    Tell it about your scene

    Plain English. One or two lines. The skill handles the rest.

  3. new chat
    JSON copy
    {
    "endpoint": "T2V",
    "aspect": "16:9",
    "duration": 10,
    "prompt_en":
    "A knight kneels in
    snow, fog rolling…"
    03

    Copy the JSON it gives back

    You get endpoint, aspect, duration, the prompt in English and Chinese. Ready to use.

  4. A knight kneeling in a snowy field — frame from the generated video.
    0:03 / 0:10
    04

    Run it through Seedance

    You get a video. That's the whole loop.

Proof

Real videos. Made with this exact skill.

Every one of these started as a comment. Pasted into the skill. Output run through Seedance. Posted.

Cinema / atmosphere
“A knight bleeding in the snow.”
UGC / selfie
“A girl filming herself in her bathroom hyping up some random serum she bought, pretending it's the best thing ever, handheld iPhone energy.”
Cinema / oner
“One uncut shot flying over a misty Japanese countryside at dawn, catching up to a bullet train, gliding through the window to a woman holding a cup of tea, then back out and pulling up into a wide aerial as it disappears into a tunnel.”
Anime / sakuga
“sakuga style animation anime fight of a fast moving samurai against a huge demon with human-like features.”

A taste of the output

Here's what it gives you.

Input was: "a defense attorney pacing in front of a jury delivering a closing argument."

seedance-prompt.json copy 
{
  "endpoint": "T2V",
  "aspect_ratio": "16:9",
  "duration": 10,
  "risk_flags": ["profession-softened"],
  "prompts": [
    {"lang":"en","prompt":"A figure in a dark suit paces steadily in front of the jury box, one hand gesturing low. \"This is where reasonable doubt lives.\" Jurors shift, one leans forward. Soft key from the jury box windows, wraps his face, amber-warm. Medium-wide handheld, 2% dolly alongside. No subtitles, no music."},
    {"lang":"zh","prompt":"深色西装的身影在陪审团席前稳步踱步，一手低位示意。\"This is where reasonable doubt lives.\" 陪审员挪动身体，一人前倾。侧窗柔光自陪审团方向来，暖琥珀色笼罩面部。中远景手持，2% 跟拍横移。无字幕，无音乐。"}
  ]
}

Speech payload

The scene depends on what's said. The line is on the page.

Three-layer lighting

Source. Behavior. Tone. Twelve words. Never more.

Risk flag

"Attorney" softened to "figure in a dark suit" — content filters love this.

●  The whole thing, free

Take the skill we use every day.

Free. No spam. Just the file in your inbox.

See the videos