Video & Animation

Direct AI-generated video that looks filmed, not generated: real people in real places, believable motion, consistent faces across shots, native dialogue, and a finished edit with captions, sound and platform masters. Covers the whole pipeline: brief, shot-by-shot storyboard, locked stills, image-to-video takes, review, edit (Remotion or ffmpeg), voice, mastering.

Turn a written script into an AI voiceover (text to speech, multi-speaker, voice cloning) with SRT subtitles and optionally a captioned faceless video, using the `voiceover` CLI from ai-voiceover-studio. Works with free voices and no API key. Use when the user asks to voice, narrate, dub or add a voiceover/subtitles to a script or video.

Design, animate, texture, and composite Vox, Johnny Harris, and WSJ style editorial explainer motion graphics across 2D, 2.5D, and 3D using installed free tools (Blender, DaVinci Resolve Fusion, Remotion + Three.js/SVG, FFmpeg) or optional open-source GIS/vector tools.

Claude'a gerçek bir 3D animatörün kamera, ışık, kaplama, hareket ve ses bilgisini vererek Blender olmadan, her karesi kodla üretilen sinematik 3D film yaptırır (Three.js + HyperFrames, MP4 çıktısı). Yürüyen ve koşan insanlar, sürülen araçlar, fizikli objeler, gerçek ölçekli kaplamalar, ıslak zemin yansımaları ve gerçek kayıtlarla ses içerir.

>-

Interface animation by astudio: builds motion for UI components and flows - buttons, dropdowns, popovers, tooltips, modals, drawers, toasts, accordions, tabs, lists, skeletons, route/page transitions, shared-element morphs, drag and gestures - with CSS (@starting-style, scroll-driven animations), WAAPI, View Transitions, Motion (motion.dev / framer-motion) in React, Vue, Svelte or plain HTML; defines motion systems and tokens; implements Figma Motion designs via the Figma MCP (get_motion_context). Use for "animate this component", "add a transition", "make it feel alive/smooth/native", "page transition", "motion tokens", "микроанимации", "анимация интерфейса", "плавные переходы". Decides whether something should animate at all before writing code.

Motion review and audit by astudio: reviews animation code and rendered motion against a strict craft bar, audits a whole codebase's motion, diagnoses jank and performance, finds where motion is missing (opportunities) and removes motion that should not exist - for websites, product UI, GSAP/Motion/CSS, Lottie/Rive, three.js and rendered videos. Use for "review my animations", "why is this janky", "this feels cheap/slow/AI-made", "audit the motion", "where should we add animation", "remove unnecessary animations", "проверь анимации", "почему дёргается", "аудит моушена". Default to flagging; approval is earned. Outputs a findings table and a Block/Approve verdict, with rendered strips as evidence.

Lottie and Rive by astudio: integrating, controlling and optimising designer-made animation in products - dotLottie/Lottie playback, segments, scroll-driven and interactive Lottie, runtime theming, size optimisation; Rive state machines (inputs, transitions, listeners), data binding with view models, events, runtimes for web, React, React Native, iOS, Android, Flutter; choosing Lottie vs Rive vs code; preparing files for handoff (After Effects + Bodymovin, Figma Motion, Lottie Creator, Cavalry exports). Use for "Lottie", ".lottie", "Bodymovin", "Rive", ".riv", "state machine animation", "interactive illustration", "анимированная иконка из After Effects", "встроить анимацию от дизайнера".

Use when the user pastes a Pinterest, Behance, Dribbble, Vimeo, YouTube or direct video link (or drops a video file) and wants the animation recreated in After Effects. Fetches or samples the reference, measures timing and motion, rebuilds it with editable keyframes, and compares frame by frame until it matches.

Motion Agent routing and best practices for AI-assisted Adobe After Effects production. Use when Codex or another coding agent needs to plan, build, modify, QA, render, or document motion design work with Motion Agent; use as the default entry point when unsure which Motion Agent skill applies.

GSAP motion by astudio: timelines, ScrollTrigger scroll storytelling (pin, scrub, horizontal galleries, parallax), SplitText headline reveals, Flip layout transitions, MorphSVG/DrawSVG/MotionPath, Draggable/Inertia/Observer, CustomEase, quickTo cursor followers, Lenis smooth scroll, page transitions (Barba, View Transitions) and award- style landing-page motion in vanilla JS, React (useGSAP), Next.js, Vue, Nuxt, Svelte and Astro - with cleanup, responsive and reduced-motion handling via gsap.matchMedia. Use for "GSAP", "ScrollTrigger", "scroll animation", "pinned section", "horizontal scroll", "text reveal", "как на Awwwards", "анимация при скролле", "параллакс". All GSAP plugins are free (GSAP 3.13+).

Compose, arrange, mix and render game music (menu themes, exploration/battle/boss loops, town and shop themes, stingers, adaptive layers) with the TuneTrack MCP tools. Use whenever the user asks for music, a soundtrack, theme, loop, jingle, stinger or ambience for a game or any project, or mentions TuneTrack.

Translate GreenSock/GSAP animation thinking into Adobe After Effects keyframes, controller nulls, text animators, expressions, and ExtendScript. Use after AEMotion when the user asks for GSAP-like professional animation, GreenSock-style sequencing, timeline/stagger/ease/scrub/Flip/SplitText patterns in AE, or asks to convert GSAP motion ideas into AE implementation.

Edit talking-head, tutorial, explainer and Short videos with Claude Code. Word-accurate cuts from raw takes, motion graphics built as HTML + GSAP (optionally three.js) pages locked to the spoken words, sound effects synthesized in Python, honest loudness handling and frame-exact rendering. On first use it asks for the creator's name, colours, logo and preferences and remembers them for every later video. Use when the user wants to cut raw footage, remove retakes and pauses, add animated cards or chapters, add or design sound effects, fix loudness, or render a final edit.

Direct and render a SaaS product launch video (60–95s, 1080p) from code. A deterministic HTML timeline rendered frame by frame with Playwright, a cloned-voice voiceover, ducked music, product-UI callouts and a macOS dock intro. Use for launch videos, product demos, Product Hunt or ad cuts.

Run immediately after the user confirms an After Effects piece is final (they approve the render or export, say it is done, or ship it). Replays every bridge operation behind the piece, distills reusable recipes, pitfalls and rules into the case library and skill/SKILL.md, and publishes them to the repository.

Modular multi-agent AI photo and video editing skill for family travel, vlogs, and unscripted highlights. Orchestrates specialized sub-agents (Client Intake, Culler & Sync, Audio/Beat Vibe Designer, Story Producer, Dual Dialogue & Montage Editors, Pluggable VFX/Color Agents, Showrunner Pacing Critic, and Executive Movie Production Director) through deterministic JSON contracts (creative_brief.json, media_index.json, beat_grid.json, paper_edit.json, pacing_audit.json, director_review.json, timeline.yaml) and renders 4K 10-bit HLG HDR deliverables + FCPXML/EDL/OTIO.

用 Codex + Seedance 从一个想法做出完整 AI 视频的八步全流程（剧情→分镜→画风→首尾帧→视频提示词→生成→后期→复核）。只要用户提到做 AI 视频、生成视频、Seedance、火山方舟出视频、分镜表、首尾帧、视频提示词、抽卡率、视频老是崩、AI 动画、短片制作，甚至只是说"想做个视频但不知道从哪开始"，都使用本 skill。也适用于想省钱地跑通 Codex 画图 + Seedance 出片工作流的任何请求。

Create and revise executable FlowSVG diagrams to explain systems, processes, and state changes with static views or animation. Validate or diagnose diagrams, manage project design preferences, and export JSON, offline HTML, SVG, PNG, or GIF. Reuse the document for playback and SDK embedding.

Draw a system architecture diagram in a technical-drawing blueprint style - services, databases, queues, caches and clients laid out automatically, with labelled request and event flows and moving data-flow dots - from a description or by reading a repository. Outputs an HTML page, an SVG and a PNG. Use when someone wants an architecture diagram, a system design drawing, a service map or a data-flow diagram.

制作或优化 3D 动画场景时，主动从用途推导视觉重点、可使用的物件细节和有因果的动态，通过 ImageGen 效果图设计确认、实际场景对照与实播验证检查完成度。适用于室内外环境、游戏空间、产品展示和叙事场景；不绑定引擎或美术风格，包含 AI 主动识别动效、专项动效检查、行为与动作分工、资产导出一致性、实播验收，以及 Blender 执行和预览脚本。

Chuyển bài viết hoặc URL thành video motion graphics dọc có kinetic typography, biểu đồ dữ liệu, voice-over tự nhiên và nhạc nền. Dùng khi người dùng muốn biến blog post, bài nghiên cứu, case study hoặc landing-page copy thành video social 9:16.

Direct a short film, kinetic type or lyric video, or a seamless README cover loop, rendered and gated.

Makes motion-graphics videos for the user — SaaS and product launch videos, feature explainers, kinetic UI demos, title sequences, end cards, HUD and text overlays — as a Claude Motion film (code-drawn, scrubbable, comment-able, exports MP4), mixing in Higgsfield clips only where real footage is needed, so the graphics cost zero credits and any detail can be changed without regenerating. Use when the user asks for a launch video, SaaS video, explainer, "motion graphics", "Apple-style", "product demo video", "animate the UI", an end card or title card, a HUD/counter overlay for a film, or a launch reel for a product.