Video Prompting Skill
AI Agent Skill for Prompting Video Models
Square-Zero-Labs · 已有 AI 审核记录
安装与使用
运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。
npx skills add Square-Zero-Labs/video-prompting-skill技能说明
Video Prompting (Agent Skill)
This is an Agent Skill for drafting and refining prompts for video generation models (text-to-video and image-to-video), and for creating character-sheet prompts for image models when the goal is consistent characters before image-to-video.
Supported video-model guides include MiniMax H3, Seedance 2.0, Seedance 2.5, LTX-2, LTX-2.3, Sora, Veo 3/3.1, Wan 2.2, Wan Animate 2, and Ovi.
The skill also includes a character-sheet workflow for:
- reusable turnaround sheets
- expression sheets and prop callouts
- scene-still handoff prompts for image-to-video pipelines
- photoreal photographic identity sheets for real-person consistency
It follows the Agent Skills spec (agentskills.io).
Tutorial
Agent Skills walkthrough and demo of this skill:
Contents
video-prompting/SKILL.md: skill entry point and workflowvideo-prompting/references/models/: model-specific prompting guidesvideo-prompting/references/workflows/character-sheets.md: character-sheet and identity-sheet workflow for image-model prompting
What It Covers
- video prompt drafting for text-to-video and image-to-video
- character-animation prompting
- model-specific routing for supported video models
- character-sheet prompting for consistent characters across longer projects
- photoreal identity-sheet prompting when the user wants real-person reference photography rather than a stylized turnaround
- recommended handoff order: character sheet -> scene still -> video prompt
Install
Claude
In Claude, install the skill using whichever format your Claude product prompts for:
- If it asks for a
.skillfile: see instructions below to build a.skillbundle. - If it asks for a folder: select
video-prompting-skill/video-prompting/.
Codex
Codex loads skills from skill directories (user-scoped or repo-scoped). Pick one:
- User-scoped (defaul