media watch references
Optimize Short Video Prompts
Research-backed workflow for optimizing AI short-video prompts and talking-head scripts. Use when Codex needs to turn a topic into traffic-oriented vertical video concepts, rewrite private-domain talking-head copy into lip-sync-friendly scripts, keep a fixed real person or reference image consistent while changing scene or emotion, or adapt prompts for tools such as Liblib/LibTV, Kling, Runway, Veo, or Sora.
Agent 安装命令
先 dry-run 再安装;仓库需已公开,或 Agent 需具备访问权限。
git clone https://github.com/yupeng0512/cursor-skills.git ~/.skillhub/yupeng-cursor-skills
cd ~/.skillhub/yupeng-cursor-skills
./install.sh --target codex --skill optimize-short-video-prompts --dry-run
./install.sh --target codex --skill optimize-short-video-prompts
~/.local/bin/skillhub-yupeng doctor --target codex --json claude
--target claude codex
--target codex cursor
--target cursor 标签
agent automation content media research