Video & Animation

Fred 的视频制作唯一日常入口。用于最终音频、逐字稿、字幕和素材到组件匹配、逐段前端建议、真实关键帧、完整 Remotion 视频及局部返修；包含正文文字动效与按需技术参考；Fred 授权时也负责按固定设计规则新做可复用组件并入库。Fred选方向、修改确认关键帧后，由助手内部检查动态、声音和接缝并交付。参考视频忠实复刻与云端执行使用各自独立入口。

Turns raw 9:16 talking-head clips into a finished, premium Instagram reel end to end: picks the cleanest take of every line across your clips, cuts dead air, punches in from 4K, cuts you out of the background for words behind your head, applies one of three designed looks (bold kinetic words, cool girl, cool dude), builds animated UI inserts with parallel sub-agents, adds subtle sound design, and checks every cut frame by frame inside the Reels safe zone. Trigger with /video-edit or phrases like 'edit these clips', 'I dropped clips in my downloads', 'cut this into a reel', 'make this reel look fire'.

Make a beat-synced product promo / launch film end to end with a human in the loop - pick music (Epidemic Sound "Create a cut"), design the storyboard together, write the shot script and an ElevenLabs voice-over script, cut the returned voice take, then build the video in Remotion (three.js metal renders, replicas of the product UI, virtual-camera push/pull/pan, every cut on the beat). Use when the user wants a promo video, trailer, launch film, product teaser, motion-graphics ad, or asks to sync video to music or generate a storyboard + VO script.

Cut, caption, and assemble vertical video with ffmpeg

>- 面向英语教师和教培机构宣传的英语批改助手视频自动化 Skill。负责环境准备、素材盘点、 差异化选题、脚本与样帧、主片生成、剪映旁白字幕混音、封面和交付整理。适用于日更短视频、 产品推广视频、QuickMarker 视频和选题池维护。

Make one branded short dev video end to end from a one-sentence idea - research, proposal, approval, voice, animated scenes, preview, render, and Buffer drafts for Instagram, TikTok and LinkedIn. Use whenever the creator gives a video idea or topic, asks for the next episode / "today's video", or says "make a video about ...".

Structure Remotion markup for interactivity

Build scroll-triggered and scroll-driven animations for websites and landing pages with React + Framer Motion (or the lightweight JOS library for non-React sites). Covers sticky scroll stacks (classic, Chronos cinematic recession, 3D tilt, multi-layer, horizontal), multi-layer parallax, immersive/Apple-style zoom heroes, gallery and product image zoom, 3D image split/shatter reveals, circular/orbital/radial/spiral motion, expanding/bell-curve cards, and canvas or video scrollytelling image sequences. Trigger this whenever the user asks to add ANY kind of scroll animation, scroll effect, scroll-triggered reveal, parallax, sticky section, or scrollytelling to a site or landing page — even if they don't name a specific effect (e.g. "make this section feel more alive on scroll", "add some motion as you scroll down"). Also trigger for simple fade/slide-in-on-scroll reveals via the JOS library.

Transcribing, displaying and animating captions

Router for all Remotion skills

Build polished, animated product overview / explainer / promo videos (MP4, 16:9 and 9:16) for a software product, from the product's REAL output (demo account, example report, dashboards, diagrams), with on-screen captions, an animated clicking cursor, UI sound effects and a music track instead of narration. Motion-graphics style like modern SaaS launch videos: dark cinematic hook scenes, light split feature scenes, animated counters, diagrams drawing themselves, a 3D browser-window pan, and an end card with the URL. Use this whenever the user asks for a product video, demo video, feature walkthrough video, promo/launch video, explainer, sizzle reel, Reel/Short/TikTok for their tool, or a video 'like the one we made for X', or short 15-second Reel/Short/TikTok cuts, even if they don't say 'motion graphics' or name this skill. Also use it to turn an existing storyboard into a rendered video, or to add music to or re-cut (e.g. vertical) a video made this way.

Make a complete faceless vertical video (30–90 s, 1080×1920) from one topic — research, script, paper cut-out images, voice-over, animation written as code (HyperFrames), render, and a second-pass review — in the "editing desk" style (one paper desk, camera flights, index cards, marker, rubber stamp, captions). Use when the user asks for a faceless video, a Reel/Short/TikTok explainer, or "make a video about X" with no footage.

Create or improve character hand-drawn music videos (手书), anime collage films, and imagegen-assisted character shorts from reference videos and character images. Use for storyboarding, per-shot paired poses, expressive motion, effects and transitions, rendering a finished MP4, or repairing awkward anatomy, malformed hands, pose ghosting, unsuitable transitions, unwanted end-card text, or stiff and repetitive character acting. Especially useful when the user requests two distinct character images per shot. Do not use for ordinary slideshow assembly or video analysis with no production intent.

Plan and produce a music video for a finished song (including Suno or other AI-generated tracks). Use when the user has an audio file and wants a music video, a TikTok/Reels/Shorts clip, a lyric-free promo video, or asks how to cut visuals to the beat.

Make a full-length music video for a song, drawn entirely from CODE (p5.js in headless Chrome: a hand-painted p5.brush cartoon, or crisp vector motion design via --style, ffmpeg for audio), with no AI image/video generation and no API spend beyond the Claude plan. Lyric-driven storyboard, phase-fitted beat grid, karaoke captions, a 30-second pilot gate, then chapters built in parallel by sub-agents. Use when asked to make, animate or storyboard a music video, lyric video or beat-synced animation, in ANY repo.

Write a storyboard and per-shot prompts for an AI video model (Runway, Kling, Veo, Sora, Luma, Hailuo, etc.) from a song's lyrics, mood and beat-synced shot list. Use when the user wants scene ideas, shot prompts or a visual concept for a music video.

Make and upload a Spotify Canvas (the 3-8 second looping vertical video behind a track) from a video clip, a music video frame or cover art. Use when the user mentions Spotify Canvas, a looping visual for Spotify, or Spotify for Artists video.

Search and understand videos without embeddings. Use when the user asks about the content of a video file or video URL - finding moments ("when does X happen"), finding a person or object from a photo, reading what is on screen at a moment, getting what is said, tracking the same things across a whole video, or one answer about the whole video. Works through the jevreel MCP tools or the `jevreel` CLI.

Make gameplay trailers for Roblox games, end to end. It plans a shot list from the game's GAME.md, drives the character in a Studio playtest so it looks hand-played (TrailerDriver.luau handles waypoints, prompts, remotes and a drifting third-person camera), adds cinematic camera moves (CameraRig.luau dolly, orbit, crane, push-in and whip, synced to game events), records only the Studio viewport with macOS screencapture, and edits with ffmpeg (speed ramps, whips, crossfades, title cards with emoji, a music bed with ducking and loudnorm). It exports the 30 s 1920x1080 Roblox video thumbnail cut and a 9:16 Shorts/TikTok cut, and runs QA with a contact sheet. Use whenever the user asks for a trailer, promo video, gameplay video, video thumbnail, teaser, TikTok or YouTube Short of a Roblox game, to "record it", "make it look like I'm playing", "edit a video of the game", or to film or capture gameplay or a cinematic flythrough in Studio.

Use when asked to record an automated demo, POC, pitch or walkthrough video of a web app (also "gravar vídeo de demonstração / apresentação de fluxo") — scripted screen recording with Playwright, on-screen captions, optional AI voice-over (TTS narration in English or Portuguese), mp4 output.

Post one video to several platforms with PostEverywhere, with the right title and caption for each. Use when the user wants to share a video, Reel, Short or TikTok across TikTok, Instagram Reels, YouTube Shorts, LinkedIn and Facebook Reels, or asks to "post this video everywhere".

Makes a launch video, trailer, teaser, demo or how-to clip of the user's own web app, as code. Recreates the app's real UI in HTML, animates it with GSAP to a music track or silent, and renders a frame-exact 1080p60 16:9 MP4 (plus a WebP for READMEs). Use when the user wants a video, promo, trailer, teaser, demo or product film of their app, site or product, a film cut to a track, a Suno prompt or generated beat for such a film, or to change or re-render an existing ship-video film. Not for editing existing footage or screen recordings, generic explainers unrelated to their app, vertical 9:16 video, or Remotion and After Effects projects.

Produz e edita vídeos realistas com IA da Voltec (provedor de internet de São João da Barra-RJ) estrelados pela personagem/modelo da marca, em Reels, Stories ou anúncio 9:16 de qualquer campanha, estação ou tema (promoção, institucional, dica de Wi-Fi, suporte, inauguração), em locais reais da região (Grussaí, Atafona, Chapéu do Sol, São João da Barra) ou em cena interna, com narração em off ou com ela falando para a câmera, portões de aprovação do cliente, snapshot aprovado antes de qualquer vídeo, ElevenLabs (gpt-image-2.5, Veo 3.1, voz clonada Radassa por TTS ou troca de voz, SFX e STT) e edição HyperFrames com legendas, CTA e QR validado; use sempre que o pedido for criar ou ajustar um vídeo, Reels, Stories ou anúncio realista com IA da Voltec ou dessa personagem (inclusive com ela falando, recortar take, trocar frase ou CTA, versão com legendas, refinar a identidade), e como base para outra marca ou persona; não use em assuntos da Voltec sem vídeo, como rede, provisionamento, GenieACS ou TR-069.

Build a talking mascot from the user's picture as-is (one pose, engine-drawn mouth and blinks) and turn a script into a lip-synced vertical video with this repo. Use when the user wants to create a mascot, add art or a voice to one, fix its mouth position, or make a mascot video from text or audio. For a full redraw with poses from reference images use draw-mascot.