video▌

The core animation hook in R3F. Runs every frame.

This skill conducts pure research for YouTube video topics. Execute all steps to produce actionable insights that identify content gaps and analyze competitors. This skill focuses ONLY on research - it does not generate titles, thumbnails, or hooks.

Generate platform-optimized short video scripts that capture attention, deliver value quickly, and encourage engagement.

Platform-specific video ad dimensions, durations, and creative frameworks for TikTok, Instagram, YouTube, Facebook, and LinkedIn. \n \n Covers exact specs for five platforms: TikTok and Instagram (9:16 vertical), YouTube (6s bumper to skippable formats), Facebook (1:1 or 4:5 square), and LinkedIn (1:1 or 16:9) \n Includes AIDA framework (Attention, Interest, Desire, Action) with timing breakdowns and hook techniques for the critical first 3 seconds \n Provides caption requirements and safe-zone

Complete guide for scripting, producing, and assembling explainer videos via inference.sh CLI. \n \n Provides three script formulas (Problem-Agitate-Solve, Before-After-Bridge, Feature Spotlight) with timing breakdowns, word counts, and pacing rules for 30–120 second videos \n Covers scene generation via video AI models, voiceover production with TTS, music ducking, and caption integration through a multi-step assembly pipeline \n Includes practical tables for pacing (120–170 wpm by content type

Perception + memory + actions for video, live streams, and desktop sessions.

使用字节跳动 Seedance-1.5-pro 模型 (doubao-seedance-1-5-pro-251215) 根据文本或图片生成视频。

This skill provides guidance for video processing tasks involving frame-level analysis, event detection, and motion tracking using computer vision libraries like OpenCV. It emphasizes verification-first approaches and guards against common pitfalls in video analysis workflows.

Core Remotion knowledge lives in .claude/skills/remotion-official/ (synced from the official remotion-dev/skills repo). This file covers toolkit-specific patterns only.

Generate talking head and avatar videos from images and audio using OmniHuman, Fabric, and PixVerse models. \n \n Four model options: OmniHuman 1.5 (multi-character), OmniHuman 1.0 (single character), Fabric 1.0 (image lipsync), and PixVerse Lipsync (highly realistic) \n Audio-driven workflow: pair portrait images with speech files to generate realistic avatar videos with synchronized lip movement \n Composable with text-to-speech and video transcription for end-to-end pipelines: generate speech