Video & Animation

Produksi video edukasi vertikal 9:16 (30–60 detik) bergaya 3D dari kode untuk seri edukasi berbahasa Indonesia, dalam tiga mode aktif — Mode A: diorama 3D semi-realistis (lanskap/sejarah, termasuk varian peta hidup); Mode B: bongkar benda (cara kerja benda/mesin); Mode C: ilustrasi anatomi 3D glossy (Seri Tubuh Manusia, ikuti-satu-aliran). Dirender dengan Remotion + Three.js; BUKAN footage asli dan BUKAN slideshow gambar AI. Mengunci timeline.json sebagai sumber kebenaran timing dari audio asli, style lock per mode, teks readout overlay DOM anti-kepotong, gerbang internal Mode Tinggal (sesi berjalan sendiri dari bahan sampai final; gerbang persetujuan klasik tersedia), QC mandiri fail-closed, render cepat per-scene, dan penutup kesimpulan/callback untuk membangun audiens.

Modular Blender 5.2.1 LTS Geometry Nodes physics workflow and isolated CLI automation across rigid_scatter, cloth_rope, and fluid_emission domains.

Logo reveal built from your mark's own geometry: grid lines and compass circles construct it while short benefit lines land beside it, then a push into its colour, your line, and a pull back to the lockup. 15.5 seconds, wide, with music; an independent style study of Android's 2019 brand film. Use when the user asks for a logo reveal, logo animation, brand opener, intro or mark animation.

Animator — make an animated explainer or educational video in Remotion from AI-drawn characters, scenes and props, with voices, music, sound effects and subtitles. Animation only, no generated footage. Use when someone wants an animation, an explainer, an educational or a "how to behave" clip.

「长卷拼贴」风格的平面动效（MG / motion graphics）设计与复刻方法：一根刻度尺贯穿的横向长卷世界、只做平移和缩放的相机、 按拍跳变的 hold 关键帧、剪影占位入场、径向/线性擦除、阈值与反相闪帧、玫红-蓝-薄荷的限色拼贴（照片抠图、线稿、 UI 截图、胶片、3D 小件）。凡是用户要做 Y2K / 拼贴 / 数据美学 / 时间轴长卷类的平面动效、要在 After Effects 里 搭这类片子、要逐帧拆解并复刻一段参考动效视频、要让 GPT/Codex 按参考帧重绘素材、或者要整理收集 AE 工程时， 都应该使用本技能，即使用户没有说出「长卷」「拼贴」这些词。Also use for: long-scroll collage motion design, frame-by-frame motion replication, AE project built from a Python scene description.

动态图形、动态排版、歌词 MV、专辑封面这类设计的判断方法，每一条都带观察依据、适用条件和代价。 内容来自对动态与静态案例的观察与复核： 画面的变化怎么对应音乐的段落、小节、重音和音色密度；换镜头时留什么不变；速度放在镜头的哪一段； 东西怎么进场、停多久、怎么退场、用什么样的曲线到位；只停一帧的闪、反相和空帧放什么、放哪里； 字的粗细、宽窄、字距、行距和几种字的搭配，字的放大、变形、裁切、遮挡与重复； 单帧的重心、留白、层次、对齐，线、点、圆、箭头、网格各自的工作；一张静帧和连续播放的成片差在哪里； 字号、歌词、画框、限色、质感、二维与三维怎么取舍；怎么逐帧拆解一段参考片、量自己的草稿、按参考复刻一段。 只要用户要设计或评审 MG、motion graphics、kinetic typography、lyric video、PV、BGA、片头、作品集、封面或海报的版式， 要让画面卡音乐，要分析或拆解参考视频，要从截图、海报或预览图推想动效，要复刻一段动效， 要选字体或处理标题字，要排一帧的构图，要定缓动和关键帧曲线，要做闪白、反相、频闪， 或者问到 After Effects 里的节奏、转场、入场、退场、停留时间、限色、小字、运动模糊、透明棋盘格， 都应该使用本技能，即使用户没有提到这些作者或这批素材。它给的是判断和核对的办法，不是可以照搬的风格模板。

Turn a project's Open Graph share card into a few-second web animation — the brand mark and title drop in from the top, then the real app rises from the bottom and settles into the final card. Builds a playable web route in the project's stack and records it frame by frame into MP4 and GIF (plus the final frame as PNG). Use when asked for an animated OG image, animated share card, social preview video/GIF, or an intro animation of the title and product.

Hacer películas y videos con código, de 15 segundos a 4 minutos, y renderizarlos a MP4. Dos modos. Historias (personajes, cómic o estilo suave, narración, música y efectos) para emocionar, saludar o invitar. Comercial (producto real o web filmada, ritmo, promesa) para demos, anuncios, Reels y videos que venden. Usa ElevenLabs para voz y música, Deepgram para oír el resultado y hojas de cuadros para verlo. Úsala cuando pidan una película, un corto, un video animado, un tráiler, un explainer, un comercial, una demo de producto, un video para Reels (vertical) o cuadrado, o «un video como el de...», aunque no nombren la skill.

Merancang storyboard pedagogis dan mengompilasi video ajar berbasis kode menggunakan framework Remotion (React, HTML5, CSS3, dan TypeScript). Menghasilkan komposisi video kuliah modular 16:9 Full HD atau 9:16 Shorts/Reels dengan animasi kinetik teks, transisi halus, syntax highlighting, diagram alir terprogram, visualisasi rumus, dan audio narasi terintegrasi.

为 Seedance 2.5 编写、改写或审查视频提示词与生成方案，处理分段、真实素材职责、角色音色、动作摄影和首尾状态；仅整理文本不授权付费生成。

Make a narrated explainer video (Opus storyboard, code-drawn frames, Polly voice, word captions, music, SFX) with swappable renderers.

Use when 将剧本拆成 AI 视频生成段，编写、精炼或审查首帧／视频提示词，处理分镜过碎、动作与切镜不清、画面重点和人物关系被操作细节淹没的问题。

Build narrative miniature worlds with a fixed orthographic camera, textured terrain, articulated people and coordinated looping actions. Use for detailed low-poly 3D dioramas or pixel-finish animated vignettes: sports grounds, street performers, gardens or other small places with a story. Four finishes, miniature, pixel, halftone and blueprint, use the same Three.js scene. This is a separate skill from the diorama SVG illustrations; prefer the diorama skill for lightweight UI illustrations, line art and empty states, and diorama-pixel for dependency-free dithered pixel-art miniatures.

An end-to-end anime production director skill for developing original anime films, episodes, shorts, trailers, characters, storyboards, animation plans, dialogue, audio, and editing workflows.

Build small animated 3D worlds for the web (isometric camera, blocks on a rounded base, soft shadows, little vehicles and people that move) in one of six themes you can tune: toy (pastel blocks, warm sun), soft, mono (one accent, the rest tints of ink), line (a hairline drawing, white fills and ink edges, by hand if asked), print (flat tones and ink, a printed illustration) and night. A hero scene for a landing page, a card-sized illustration for docs or a blog, or a mini figure for any small spot (an empty state, a feature in a list, a 404). Use this whenever someone asks for a 3D illustration, an isometric scene, a low-poly or "toy town" animation, a monochrome or line-art 3D figure, a hand-drawn or sketched figure of an object with a bit of life in it, a blank/empty-state illustration, an illustration that answers a hover or a tap, a little game-like SVG world, even if they never say "3D" or "isometric". Ships one self-contained HTML file that draws SVG with no dependency (no th...

Use when you need a short video of a page's real animation for a launch post, a LinkedIn update or a team review - or when a screen recording of a scroll-driven page is choppy, mistimed, shows things it should not, or does not read in a muted feed.

Build animated pixel-art miniatures: a small isometric place drawn at low resolution and enlarged crisp, with ordered dithering on every surface, procedural planks, thatch, paving, grass and water, a ground that dissolves into the white page, soft cast shadows and tiny people with a looping action. One self-contained HTML file with no dependency (no three.js, no WebGL): a software renderer on a 2D canvas, about 50 KB. Use it whenever someone asks for a pixel-art scene, a retro or 16-bit isometric vignette, a dithered illustration, or a cozy little place in pixels for a landing page, a blog post or a card. For vector UI illustrations, line art and empty states use the diorama skill; for detailed Three.js worlds with articulated people and four finishes use diorama-worlds.

Escolhe a moldura do vídeo entre tela dividida, rosto com janela, tela com rosto no canto, tutorial mudo em loop e rosto com legenda. Use quando ele quiser outro formato de vídeo, testar versões do mesmo conteúdo ou fazer um tutorial sem fala.

Produz um Reels do cliente com roteiro e storyboard aprovados antes de gravar, animações presas à fala, sincronização com a gravação e render final. Use quando ele pedir um vídeo, Reels ou roteiro novo, ou para continuar um vídeo já roteirizado (por exemplo "/fabrica-videos-reels <tema>" ou "gravei o <slug>").

制作无脸说明视频的小红书 3:4 截图卡短视频：把真实 X 截图、动态背景和已核验本地 BGM 组装成一个视频与一份发布稿，并按周归档；不自动发布。

创作者的主力 AI 视频提示词 Skill。先按脚本风格选档位，再像导演一样把脚本、故事、口播稿、分镜表或一句想法，拆成能直接生成的分段视频提示词。短视频档（抖音/小红书节奏：一镜到底、多分镜、街拍、角色小故事、竖屏短剧，标杆是《丹眉》和赛博原子）和电影精品档（慢节奏剧情、表演、对白、对手戏、AU 微表情、八维公式），各有结构和示例。每段含生成模式与参数、参考图绑定、人物道具场景、镜头时间轴、声音和硬约束。适用 Seedance 2.0/2.5、即梦、dreamina、LibTV、TT AI。Use when 用户要写视频提示词、AI 视频提示词、Seedance 提示词、即梦提示词、脚本转视频、拆分镜、分镜提示词、快剪、短视频提示词、storyboard、shot list、多图参考、首尾帧、人物表演提示词、AI 演技、情绪递进、导演式提示词，或给出脚本/故事/想法要变成视频。当前项目若有专用工作流 Skill（Vlog、情感钩子、新闻解说等），优先用专用的。

把文字、文章、笔记、PPT 或 HTML 课件提炼为有配音字幕的 16:9 知识讲解视频。 使用 Remotion 语义动画与 Fish Audio / ElevenLabs 配音，先做短样片再扩展。 用于课件转视频、知识视频、内训片、科普 explainer，也用于参考视频的解释型动效拆解与原创复现；不负责真人口播剪辑、数字人生成或未知作者工具归属鉴定。

动效 KINETIC — 把口播视频（数字人或真人实拍）做成爆款动效质感的视觉语言与逐帧渲染器。用一条贯穿全片的"光"、按音节落下的大字、剪辑师才认得的圈内物件（框选框、真实声波、关键帧、时间线、口型条）、人物抠像后"字藏身后"、音画同钟的合成音效与节拍对齐的配乐。用户说"用动效做/用动效风格/做得更炫/像那条爆款动效视频一样/Opus 那种动效"、或要给数字人/真人口播加高密度动效包装时使用。不用于：只要简单字幕的口播、纯 MG 无人物讲解（用 knowledge-explainer）、结构化信息单镜头（shotcraft-motion）。

Animate characters, creatures and props in 2D and 3D in any style: the 12 principles, timing and spacing, keys and breakdowns, graph editor curves, cycles, IK or FK per shot, shape key and face swaps, frame by frame and cut-out, and playblast checks. Use for character animation, walk cycle, run cycle, idle loop, squash and stretch, anticipation, overshoot, on twos, Dope Sheet, Graph Editor, Grease Pencil animation, Krita animation, cut-out animation (\"walk cycle\").