seedance-2-prompt

Generates 3-15 second video clips via Seedance 2.0 through OpenRouter or Volcengine ARK backends.

6.8k|540|Updated May 6, 2026
One-click install
npx skills add https://github.com/opensquilla/opensquilla --skill seedance-2-prompt
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: seedance-2-prompt
Source: https://github.com/opensquilla/opensquilla/tree/main/src/opensquilla/skills/bundled/seedance-2-prompt
Command: npx skills add https://github.com/opensquilla/opensquilla --skill seedance-2-prompt

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires httpx, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Creating short AI-generated video clips requires navigating different provider APIs, authentication schemes, and polling workflows. This Skill submits a structured prompt to Seedance 2.0, polls the generation job, downloads the resulting MP4, and verifies it with ffprobe, all through one unified command.

Core Features & Use Cases

  • Dual-backend video generation: Submit jobs to OpenRouter (bytedance/seedance-2.0) or the official Volcengine ARK / BytePlus ModelArk endpoints with the same CLI interface.
  • First-frame and reference image control: Lock the opening frame with an input image or anchor character identity and style across shots with reference images.
  • Verified output with receipts: Downloads are validated with ffprobe, atomically published, and accompanied by a sanitized JSON receipt recording provider, model, and verification metadata.
  • Use Case: Produce a 10-second vertical cinematic shot of an original character for a storyboard, then generate additional segments and stitch them with a video-merger workflow for content longer than 15 seconds.

Quick Start

Ask the agent to generate a 5-second 9:16 video of a young woman in a red trench coat walking through rain-soaked neon streets and save it as scene1.mp4.

Frequently Asked Questions about seedance-2-prompt

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate an AI video clip with Seedance 2.0?

Provide a structured English prompt and an output filename, then the script submits a job to the Seedance 2.0 API, polls until completion, and downloads the MP4. The default provider is OpenRouter using the bytedance/seedance-2.0 model.

What is the difference between OpenRouter and Volcengine ARK for Seedance video generation?

OpenRouter uses a flat prompt field and returns a polling URL with the video in unsigned_urls, while Volcengine ARK and BytePlus pack inputs into a content array and return a job id polled at a constructed task URL. Both are normalized into one CLI contract selected with the provider option.

Can I keep the same character consistent across multiple video shots?

Yes, pass the same reference image as input_reference to every shot to anchor identity and style without locking the frame. Use input_image only when you need a specific literal first frame, and repeat the character's full description in each prompt.

What video durations and aspect ratios does Seedance 2.0 support?

Durations range from 3 to 15 seconds, with 3-second requests generated as 4 seconds and trimmed locally. Supported aspect ratios are 9:16, 16:9, 1:1, 4:3, 3:4, and 21:9, with resolutions of 480p, 720p, or 1080p on the Volcengine path.

How do I create videos longer than 15 seconds with Seedance?

Generate individual segments of 15 seconds or less, ending each on a stable handoff frame, then extend the previous output or stitch segments with a video-merger skill. Each segment should carry forward identity, outfit, lighting, and camera style.

Why did my Seedance video generation fail with a policy rejection?

Provider policy rejections typically come from franchise names, brand terms, or recognizable character likenesses in the prompt or reference images. The script stops immediately, writes a sanitized failure sidecar with the policy code, and never retries a paid submission automatically.