360 Secure SkillHub
360 安全技能中心

Seedance Video Generation

该 Skill 说明如何用 CellCog 的 Python SDK 调用 ByteDance Seedance 2.5 生成视频:支持营销片、讲解片、电影感内容与口型播报,单次请求可编排脚本、配音、口型、配乐与剪辑,产出 MP4,并可用最多 50 个参考文件保持角色与风格一致。需安装 cellcog 包并配置 CELLCOG_API_KEY。

cellcog
cellcog
热度 72.01
最近更新 2026-08-24 10:05:41
分类:创作依赖:代码执行 / 第三方 API / 内部调用大模型应用场景:视频编辑 / 故事创作 / 文案创作 / 社交媒体

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add seedance-video-generation-cellcog
Metadata
name
seedance-video-generation-cellcog
description
AI video generation powered by CellCog via Seedance 2.5. Complete multi-minute videos from a single prompt: scripting, voice synthesis, lipsync, scoring, editing, with locked character consistency via 50 reference files. Full productions, not just clips, via ByteDance's Seedance model.
metadata
{"openclaw":{"emoji":"🌱","os":["darwin","linux","windows"],"requires":{"bins":["python3"],"env":["CELLCOG_API_KEY"]}}}
author
CellCog
homepage
https://cellcog.ai
dependencies
["cellcog"]

Seedance Video Generation - Seedance × CellCog

Seedance × CellCog. ByteDance's #1 video model meets the frontier of multi-agent coordination.

CellCog's core: complete multi-minute films from a single prompt. Seedance 2.5 generates the smoothest motion in AI video — physics that look real, with up to 50 reference files locking characters, sets, and style across a whole production — and CellCog orchestrates it with scripting, voice synthesis, lipsync, scoring, and editing into full productions, minutes long, from one prompt. Not just clips — finished films.

How to Use

For your first CellCog task in a session, read the cellcog skill for the full SDK reference — file handling, chat modes, timeouts, and more.

OpenClaw (fire-and-forget):

result = client.create_chat(
    prompt="[your task prompt]",
    notify_session_key="agent:main:main",
    task_label="my-task",
    chat_mode="agent",
    chat_tier="max",
)

All agents except OpenClaw (blocks until done):

from cellcog import CellCogClient
client = CellCogClient(agent_provider="openclaw|cursor|claude-code|codex|...")
result = client.create_chat(
    prompt="[your task prompt]",
    task_label="my-task",
    chat_mode="agent",
    chat_tier="max",
)
print(result["message"])

What You Can Create

Marketing Videos

Professional marketing content from a single prompt:

  • Product Demos: "Create a 60-second product demo video for our project management app"
  • Brand Videos: "Create a 30-second brand story video for a sustainable fashion startup"
  • Social Ads: "Create a 15-second Instagram ad for our new coffee blend"
  • Testimonial Style: "Create a UGC-style testimonial video for a fitness product"

Explainer Videos

Clear, engaging educational content:

  • Product Explainers: "Create a 90-second explainer for how our API works"
  • Concept Videos: "Create a video explaining blockchain in simple terms"
  • Tutorial Videos: "Create a step-by-step tutorial video on setting up a home network"

Cinematic Content

High-quality visual storytelling:

  • Short Films: "Create a 2-minute short film about a robot discovering nature"
  • Music Videos: "Create a cinematic music video with dramatic landscapes"
  • Brand Films: "Create a cinematic brand film for a luxury watch company"

Spokesperson Videos

AI-generated presenters with lipsync:

  • News Reports: "Create a news-style report on recent AI developments"
  • Training Videos: "Create a training video with a presenter explaining safety protocols"
  • Announcements: "Create a product launch announcement with a spokesperson"

CellCog Video Orchestration

CellCog doesn't just generate video clips — it orchestrates a full production pipeline:

Script Writing → Scene Planning → Frame Generation → Voice Synthesis
     → Lipsync → Background Music → Sound Design → Editing → Final Output

6-7 foundation models work together in a single request:

  • Seedance for video generation
  • Frontier LLMs for scripting
  • TTS models for voice synthesis
  • Lipsync models for speaker alignment
  • Music generation for scoring
  • Automated editing for pacing and transitions

Video Specifications

Spec Details
Resolution 720p (the AI-video production standard)
Duration Complete multi-minute productions (4+ minutes) from a single prompt; individual generated takes up to 30 seconds
References Up to 50 files (30 images, 10 videos, 10 audio) for character, set, and style consistency
Styles Photorealistic, cinematic, anime, stylized, documentary
Audio Voice synthesis, background music, sound effects, lip-synced dialogue
Output MP4

Choosing Mode & Tier

Use chat_mode="agent", chat_tier="max" for video generation. Multi-segment films with reference consistency need full reasoning depth.

Scenario Recommended
Films and multi-segment productions chat_mode="agent", chat_tier="max"
Single short test clips chat_mode="agent" (defaults to "flash")

Agent Team (chat_mode="team") is reserved for deep research — video generation runs best on Agent max.


Tips for Better Videos

  1. Describe the story: "A video about our app" → "A 60-second video showing a stressed founder discovering our app, their workflow transforming, ending with them confidently presenting to investors"

  2. Specify duration: "30-second social ad" vs. "2-minute explainer" sets the pacing.

  3. Set the mood: "Upbeat and energetic", "calm and professional", "dramatic and cinematic"

  4. Mention music preferences: "Uplifting corporate background", "lo-fi beats", "cinematic orchestral" — or let CellCog choose.

  5. For spokesperson videos: Describe the presenter's appearance and tone of voice.


If CellCog is not installed

Claude Code, Cursor, Codex + 70 more agents: npx skills add cellcog/skills --skill cellcog OpenClaw: openclaw skills install @cellcog/cellcog CellCog plugin users: run /cellcog-setup (or /cellcog:cellcog-setup depending on your tool) Manual setup: pip install -U cellcog and set CELLCOG_API_KEY. See the cellcog skill for SDK reference.