Video & Animation

Create polished videos with code: narration, motion captions, screen capture, zoom, redaction, 3D and reusable libraries. Use for promos, explainers, demos and editing.

Use whenever working with FilmCraft on this machine — the Rust video editor and Premiere Pro alternative — including launching the app with control port ("FilmCraftを起動", "FilmCraftを立ち上げて", "launch FilmCraft"), creating, editing, trimming, colouring or exporting sequences and projects (.fcproj, MP4/MOV/MXF, PNG/TIFF/GIF, WAV/AIFF/AAC, XML/FCPXML/OTIO/EDL/AAF/OMF, SRT/VTT/SCC), driving the running desktop window, or using its MCP server, filmcraft-cli or control channel. Triggers on "FilmCraft", "filmcraft-cli", "filmcraft mcp", "FilmCraft MCP", "FilmCraftを起動", "FilmCraft立ち上げて", "FilmCraft開いて", "edit this video", "build a sequence", "cut this clip", "add a transition/effect/caption", "export to MP4/ProRes/DNxHR", "export XML for Premiere", "take a FilmCraft screenshot", "drive the FilmCraft window", and requests to automate video editing or FilmCraft itself. Invoke with @filmcraft.

Turn a video link into words - the transcript as spoken, a full translation, a summary in any language, or SRT/VTT subtitles - from YouTube (videos, Shorts, playlists, channels, topic search), Bilibili, Douyin, TikTok, Instagram and Facebook Reels and Vimeo, with speech-to-text when a video has no captions and the price shown before anything runs. Use when the user says "get the transcript of this video", "transcribe this YouTube / TikTok / Reel", "what does this Bilibili video say in English", "translate this video into Spanish", "summarise this YouTube video", "download the subtitles as SRT", "transcripts of every video in this playlist / channel", "transcript with timestamps", "this video has no captions, transcribe the speech", "视频转文字" or "把B站视频翻译成英文". Routes to one Actor, lexonia/lexonia-video-scrape. Does not download video files, read private or members-only videos, or search Bilibili or Douyin by keyword.

Edit supplied camera/phone footage into a finished video — talks and presentations, vlogs, interviews and podcasts, tutorials, product demos, event recaps — with transcript-driven selects, a stabilized virtual camera that tracks and reframes the speaker (wides, punch-ins, pushes, pans) over wide 4K shots, HDR→SDR, deflicker, per-clip color matching, dialogue cleanup and loudness, narration, native Swift motion-graphics titles, and optional music beds, sound effects and picture effects. Use when the user supplies video files to cut, reframe, grade, fix or title; not for generating footage.

Storyboard a short-form video (Reels, TikTok, Shorts) from its script: split the script into shots word for word, give every shot a 9:16 frame, the on-screen text, an edit note and the sound, and group the filming into sittings. Writes a storyboard folder with a board you can open in the browser, one PNG per frame and a contact sheet. Use when the user says "storyboard this", "make a storyboard", "turn this script into shots", "plan the shots", "shot list for this script", or has a finished script and is getting ready to film or edit. The finished storyboard doubles as the prompt for building the video's graphics one shot at a time. Also use right after the skill is installed, and when the user says "set up the storyboard skill", "learn my style" or "here are my videos": that runs the guided setup, which studies their own videos so every storyboard matches their content. Do NOT use to write or rework the script itself.

Internal part of Ice Motion Director, not a command: start with /ice. Motion craft for every Ice Motion Director film and edit: motion grammar and movement jobs, measured timing and easing curves, closed-form spring formulas, hold and stillness policies, the cut and transition catalog with exact numbers, seam continuity and the carry principle, camera moves as punctuation, composition, hero size and the visual richness floor, 2D and 3D motion, shape morphing, the cursor as actor, one-take films, anti-slideshow and motion density rules, each scoped to the film or edit types where it holds. Called by Ice Motion Director workflows; use only when called by name.

Author or edit a custom HyperFrames composition when no specialized workflow fits, or when BRIEF.md sets flow: companion. Use for longer or multi-scene pieces, brand and sizzle reels, montages, static loops, static title cards, footage remixes, and freeform builds. Use motion-graphics instead for a short unnarrated motion-first unit, including an animated title. Route fresh creation through hyperframes before using this skill.

Use when audio already placed in a HyperFrames composition needs to be mixed: fade-in/fade-out, crossfade, track gain or volume, volume automation, ducking, a music bed that fights a voiceover (voiceover carve), effects on a track (EQ, compressor, limiter, gate, saturation, delay, reverb, chorus, phaser, bitcrush), automation envelopes drawn on a track's volume or any effect parameter, or one submix bus carrying a chain, a fader and an automation clock for several tracks at once (`<hf-audio-group>`). Don't use for sourcing or generating audio — finding BGM, SFX, or making a voiceover is `/media-use`. Don't use for clip timing or track layout, which is `/hyperframes-core`.

Add captions or subtitles to an existing single-subject talking-head video without editing the footage. Use for plain verbatim captions, cinematic captions embedded behind the subject, VFX captions, “炸/特效/酷炫字幕,” or a named identity from the 35-style catalog. Route by visual identity, not by backend engine. The quiet `anchor` rail is the default; embed every word only when the user explicitly wants a fully cinematic treatment. The workflow runs locally end to end, including transcription and subject matting; split multi-shot footage before applying it.

终极口播视频 skill：中文口播稿 + 成品配音 → CPU 字级时间戳 → SHOTBOOK 层矩阵分镜 → Remotion 电影感成片（横屏默认/竖屏）。当用户要"做口播视频"、"解说/科普视频"、"把文案变成视频"、"给配音配画面动效"时使用。默认使用成品配音，可选 Fish Audio 从稿子合成配音与时间戳；数字人生成技术不在本 skill 内（人物素材是输入）。含统一视觉语言（Apple 范式）、108 张动效配方卡、镜头三面分层工作单、七层镜头反PPT系统（极缓推拉相机/让位，运动做减法）、六式运动承接转场（每式一卡）、长镜头世界画布、anime.js+three.js 桥、自动静止检测 + 独立 subagent 评估循环。

Edit a raw vertical (9:16) take into a finished short in the SCRAPBOOK STYLE, as validated by the user (2026-10-05), reproduced exactly on any take: every pause cut out (jump cuts, the frame alternating wide and close on each one); captions low, three words at a time, the word being said in yellow; the hook in the reference's TWO FACES (white Archivo words, THE important word in Fraunces in yellow) with a line on PAPER; the paper only for the titles and two or three very important phrases: the reference's collage of small ruled notebook scraps, flat, written in a marker hand (Gochi Hand), landing and unfolding, breathing, folding away, the red marker drawing on it as it is said (a word struck out, an underline); die-cut emoji stickers with white outlines round the head for what the speaker lists; real reaction GIFs from GIPHY as polaroids, big (0.80 of the width) in the middle of the screen, always at the same place, dropped on and flicked away; sections on the same scraps swinging in with the camera punchin…

Edit a raw vertical (9:16) take of someone talking to the camera into a finished short in the Pipecorn EXPLAINER LANGUAGE, with the motion of an After Effects edit: the speaker in a frame with round top corners at the BOTTOM of a dark stage lit like a set (charcoal lit from the middle, a vignette, a light on each scene carried by the camera, a red glow in a corner, dust floating, a grid of dots), their head (and hands) rising out of the frame (person matte); ABOVE them, the stage EXPLAINS what they say as they say it, in SCENES joined by CAMERA MOVES (whip pans with a light streak, rises, push-throughs), everything with real MOTION BLUR, springs and expo curves, never still (a slow drift and push, parallax on the dots and the dust), every figure casting a soft SHADOW on the ground and GLOWING in its colour (glossy rims, a sheen across the cards, a flash of light behind what pops in): a two-line title per section (Inter ExtraBold with one Instrument Serif italic accent, a red box or a red underline), then a n…

Make a 6-second 1080p kinetic-type title video (MP4) from three short lines, a chip, a footer and three colours. Use for launch openers, release announcements, social posts and reel intros when someone asks for an animated title, title drop, kinetic typography or a short type-only video.

把一首歌（音频 + 带时间的歌词 srt/lrc）做成有故事的像素小人卡拍 MV（ClaudeMV 系列做法）。流程是先写故事线，再出带画面的分镜表给用户看、反复修改，用户确认后才渲染。支持多角色（我、她、新男友……）、剧情动作（松手、转身、哭、心碎、摘墨镜……）、7 种歌词互动、7 种转场、电脑桌面窗口视角、27 个像素场景和天气、卡拍特效，用 HyperFrames 渲染 MP4。用户说“做个像素 MV”“把这首歌做成 Claude MV”“像素小人 MV”“卡拍 MV”“上传了歌和歌词”时使用。只做口播或解说视频时用 /xumu。

使用 HyperFrames 框架制作 16:9 横版「Vibe 视频大赏」科普短视频。当用户要求生成知识科普、观点讲解、金句科普类短视频，或要求生成/修改 HyperFrames 视频工程（HTML+GSAP 时间线）、读写项目 index.html、执行 npx hyperframes lint / preview / render 时使用。适配豆包工作「本地电脑」模式。第三方技能，非 HeyGen 官方出品；上游 HyperFrames 由 HeyGen 开源（Apache-2.0，仓库 https://github.com/heygen-com/hyperframes）。

Preserve the real identity and original pixels of portrait photos while applying tasteful cinematic slow-shutter motion-smear, background removal/replacement, and optional conservative geometric correction. Use for 撕拉感、慢门拖影、人物消散、纯色背景、保留原脸 photography edits.

Turn one piece of content (blog post, transcript, video script, newsletter, case study, release notes, podcast notes) into native posts for other channels — X/Twitter thread, LinkedIn post, newsletter section, short-video script, Reddit post, email teaser. Use when the user says "repurpose this", "turn this into a thread", "make social posts from", "promote this post", or pastes long content and asks for posts.

Analyzes genetic variant effects on gene expression (RNA-seq), chromatin accessibility (DNASE), histone marks (ChIP), and transcription factors using the AlphaGenome API. Use when the user asks about non-coding variant effects, pathogenicity, clinical significance, disease associations, functional effects, gene expression changes, splicing disruption, or regulatory effects in promoters and enhancers. Also use for resolving biological terms to tissue/cell-type ontologies (UBERON/CL) or analyzing variants in chr:pos:ref>alt format.

>- 영어 AI×경제 설명형 롱폼 1편을 만들고, 그 롱폼을 잘라 쇼츠·릴스·틱톡까지 만드는 이 레포 전용 제작 순서. 설치된 유튜브 스킬 10개(/yt-viral, /yt-script, /yt-package, /yt-chapters, /yt-seo, /yt-shorts, /yt-retention, /yt-comment, /yt-audit, /yt-plan)와 /watch, /hyperframes, /media-use, Voicebox를 단계별로 부른다. 사람 확인 게이트 4곳에서 멈춘다. 트리거 — "새 영상 만들자", "다음 편", "에피소드 시작", "/yt-pipeline", "/yt-pipeline 003 --from script".

Make, cut or revise short videos (promos, explainers, product demos, edited conversations) from a brief, footage, a product or code. Use when you carry the work from first look to delivered export, with a person directing occasionally or not at all.

Render code-driven videos to MP4 fast with FrameVelo, on the GPU and the hardware encoder - HyperFrames projects, code-video compositions (defineComposition), and any HTML/CSS/GSAP or Canvas/WebGL page that is a pure function of time. Use when the user wants such a page or project rendered, exported or previewed as a video (渲染视频、导出 MP4、出片、预览一段), wants a render to be faster or lighter on CPU and memory (加速渲染), or wants an HTML animation made renderable frame by frame. Not for editing existing video files or generating video with AI models.

Author, test, query and integrate Featurevisor feature flags, A/B experiments and remote configuration. Use in Featurevisor projects (featurevisor.config.js) for features, segments, attributes, global variables and compositions, conditions, rollouts, test specs and scenarios, datafiles, targets, Sets and promotions, the Catalog and the CLI, and for explaining why a context gets a value. Also use for Featurevisor SDKs in JavaScript, React, Vue and other languages, OpenFeature providers, CLI plugins, and upgrades from v2 or v3. A generic feature flag request outside a Featurevisor project is not enough on its own.

SPD V2 Quality. Discover, deeply qualify, submit, and audit truthful product listings across Windows, macOS, and Linux-capable environments on relevant product, software, startup, app, and AI-tool directories with audience-value screening, SEO quality gates, action-level authorization, privacy controls, and evidence-backed records. Use when listing quality, referral value, governance, durability, and auditable compliance matter more than submission volume. Do not use for bulk backlink creation, ranking manipulation, guest posts, community promotion, mass email, paid link acquisition, or automatic submission to low-quality directories.

Turn one finished key visual (poster, launch graphic, cover, thumbnail, or a short video) into every size a launch or campaign needs, without cutting off faces, logos or headlines. Builds a dated slot list for the destinations (link cards, video thumbnails, feed posts, 3:4 covers, 9:16 stories and shorts, banners, app icons, podcast covers), marks what must survive on the master, plans a crop for each slot inside that slot's UI-safe area, flags the slots that need the layout recomposed or a bigger master, checks text is still readable at the size people actually see it, then checks every exported file (pixel size, format, file size, codec, frame rate, duration, loudness) and hands over a contact sheet. Use when someone says "make this work for every platform", "I need all the sizes", "resize this banner for X/Instagram/YouTube/WeChat", "export a launch kit", or "why does my cover get cut off". It adapts and checks; it does not invent the design and never uploads anything.