Video & Animation

AI 口播配音：用免费的 edge-tts（微软 Edge 在线语音）为带货短视频逐句生成配音，自动去首尾静音、统计每句时长、缓存复用，可选音色/语速/音调，并给出口播稿改写规则。Use when the user needs a voice-over, 配音, 口播, 旁白, TTS, text-to-speech, AI 主播声音, or wants to turn a script into audio for 抖音 / TikTok / 视频号 videos.

ffmpeg 自动拼接成片：一条命令把 storyboard.json（商品图/视频片段 + 口播稿）渲染成带配音、字幕花字、转场、BGM 自动闪避、响度标准化的 9:16 MP4，纯本地运行。Use when the user wants to assemble/edit/render a product video, 自动剪辑, 一键成片, 拼接视频, 图片转视频, Ken Burns, add BGM/subtitles/voice-over with ffmpeg, or batch-produce short videos without an editor.

封面与标题 A/B：为带货短视频生成多套封面（自动挑选清晰帧或用干净素材图 + 不同标题与版式）和多组标题/文案，导出 3:4 / 9:16 封面文件和对比图，并给出 A/B 测试方法与记录表。Use when the user wants a video cover, 封面, 首图, thumbnail, 标题, 文案, title variants, A/B test covers for 抖音 / 视频号 / 小红书 / TikTok, or wants to improve click-through rate.

Records a narrated walkthrough video (MP4 with voice-over and captions) of a web app by driving a real browser. Use when someone asks for a demo video, product walkthrough, feature tour or screen recording to show a reviewer, client or team. Works on Windows and macOS.

影片加強版的原理解說動畫流程（HyperFrames + 中文旁白 + 字幕 + 三色主題與字型 + 授權照片與 Codex 生成圖 + MiniMax H3 AI 生成影片 + 自我審查），產出可直接上課投影或上傳 YouTube 的 MP4。影片素材以 AI 生成為主，避開版權問題。使用者要做解說動畫，並且想加入影片、AI 生成影片、H3 影片、動態實景畫面時，就一定要使用此技能。觸發情境包含：「做一個有影片的解說動畫」、「用 H3 生成影片搭配講解」、「動畫裡加 AI 影片」、「explainer video with AI footage」。

動畫影片的視覺自我審查循環（HyperFrames 版）— 輸出草稿、在每個時間軸節拍截圖拼成 contact sheet、逐張對照檢查表、修正、重來。只要動畫的畫面或時間有任何修改，以及在宣告影片完成之前，都一定要使用此技能。

Create Russian educational motion-design videos for the user's YouTube Shorts and Instagram Reels channel, especially requests for another topic in the approved React-video style, with Dmitry narration, music and the supplied frog subscription outro.

動態設計品味規則（HyperFrames 版）— 時間軸、緩動、彈性、節奏、字體（含繁體中文）、配色、質感與場景連貫，全部附具體數字。只要是製作、調整、重新計時或打磨動畫影片、片頭、轉場、motion graphics、HyperFrames composition，就一定要使用此技能。觸發情境包含：「做一支動畫」、「做動態影片」、「做片頭」、「motion design」、「讓這段字動起來」、「調整動畫節奏」等。

Make a broadcast-quality promo video (MP4, 16:9 / 9:16 / 1:1) for any website, app or product, rendered entirely in code around one continuous light - six film formats (journey, keynote, launch, story, showcase, reel) so every product gets a different shape, real or rebuilt UI, kinetic type, device loops, product shots, a voice-over, an original per-product score and full sound design. Use when the user asks for a promo, product video, launch film, ad, spec ad, demo video, teaser, reel or brand animation for a URL, an app, a product or an idea, in a light or dark look.

用代码做配旁白的短视频，不用视频生成模型。适合以图形、文字、代码和数据图为主的视频，例如知识讲解、软件教程、产品功能介绍、数据解读、清单盘点；不拍摄、不生成写实画面，所以不适合实拍、真人出镜、剧情和实物展示。做法：Canvas 按时间画出每一帧，edge-tts 配音并按逐词时间戳对齐画面节拍，Playwright 驱动本机 Chrome 逐帧截图，ffmpeg 合成，再出 16:9、4:3、3:4 三种封面和联络表。适用于新建这样一个视频项目、新增一期、改旁白或画面、出静帧检查画面溢出。

每天定时从抖音创作者中心、视频号助手、小红书创作服务平台只读采集作品数据（播放、完播率、2 秒跳出或快速划走、平均播放时长、点赞收藏转发评论、涨粉、流量来源、搜索词、留存曲线），存成按日期命名的快照，再排成每条作品的时间序列。采集时间写在配置文件里。适用于设置每日采集、手动补采、看某条视频的数据走势、写周报前汇总数据。

用 HyperFrames 生成 1080×1920 竖屏双语歌词视频：深色梦幻背景 + 单个透明玻璃几何体撞边变形 + 底部小动物乐队 + 居中「原文 + 中文」字幕，时间戳来自 Whisper 实测。当用户给出歌曲音频（MP3/WAV）和歌手名字，要求做歌词视频、歌词 MV、带字幕的音乐视频、给一首歌配歌词动画时使用。音频路径与歌手名是唯二必需输入；用户另给官方歌词时优先采用（歌词为可选项，不给就用 Whisper 重建）。

发布短视频到抖音、视频号、小红书（B 站只写文案）：先读抖音、小红书和 Google 的搜索联想词，写标题、描述、话题、封面字和置顶评论；再用 playwright-cli 在用户已登录的 Chrome 里代填发布表单，上传视频和封面，填合集、位置、AI 声明（用了 AI 时）、热点和活动。不管视频是怎么做的都能用。默认停在发布按钮前，用户明确说发布这一条之后才点。适用于写或改发布文案、查联想词、选话题、代发视频、换视频或封面、核对已填好的表单。

Autonomous director for code-drawn cinematic music videos. Analyzes a song, aligns lyrics honestly, develops and selects an original story, builds a frame-accurate shot plan, draws a deterministic 2D/ink animation in headless Chrome, renders shot-resumably, muxes the original audio, and runs technical plus visual QC. Use for narrative MVs, animated short films, and music-led title sequences.

Bilinen Türk meme kliplerini (kendi orijinal sesleriyle), Türkçe TTS anlatımı, kelime kelime altyazıyı ve veriyle beslenen animasyon sahnelerini birleştirip 1080x1920 dikey Instagram Reels / TikTok / YouTube Shorts videosu (MP4) üretir. Kullanıcı bir ürünü, özelliği, lansmanı, hatayı ya da hikâyeyi "reels", "reel", "kısa video", "tiktok", "shorts", "meme'li video", "dikey video", "sosyal medya videosu" diye anlatmak istediğinde, ya da daha önce üretilmiş bir meme reel'in benzerini istediğinde bu skill'i kullan; kullanıcı "skill" demese de.

将本地AI兵棋对局制作成有真实战报依据的视频：独立编队指挥、双盲与通信、逐回合监督、选局、机制解说、观众互动、剪辑和封面。适用于海战及其他兵棋；可只执行用户要求的一环，不用于把虚构战况包装成实战。

Create or edit local videos with Remotion and HyperFrames. Turn audio plus untimed MD/TXT lyrics or timestamped captions into a verified lyric MV using the user's continuous scrolling, glyph-masked progress style. Also retain the integrated general video workflow.

>- Turns a plan, research note, proposal, roadmap or decision list into a narrated explainer video (MP4) that walks through why it matters, how it works, and each open decision with its options. A thin wrapper over the HyperFrames /hyperframes and /faceless-explainer workflow that fixes the brief, the story shape and a shared "question stage" layout, and patches the steps that commonly break on a fresh machine (local Kokoro voice, brand fonts, timing). Use it whenever the user wants to explain a plan, a set of decisions, open questions or research findings as a video, even if they do not say "explainer" or name HyperFrames. Triggers: "turn this plan into a video", "explain this as a video", "make a video of this research", "decision video", "walk the team through these options in a video". Not for product promos from a website URL, PR walkthroughs, captions on existing footage, or logo stings.

Burn on-screen text onto a video so it looks exactly like TikTok's own in-app text tool (TikTok Sans "Classic" font; plain white, white with black outline, pale yellow, gold yellow, lilac; Apple iPhone emoji) - either copying the captions of a reference TikTok frame-accurately (wording, line breaks, size, position, timing) or writing new captions. Use whenever the user asks to add text, captions, an overlay or a hook to a video, to copy/recreate a reference video's text, or says "with text".

Agent 2 of the video editing pipeline. Finds and removes spoken mistakes — stutters, repeated words, false starts, and retakes (re-recorded lines) — from a talking-head recording. Use after cut-silences, when asked to cut mistakes, remove stutters / repeats / filler restarts, clean up flubs, or keep the best take. Works review-gated: it proposes every cut with context and a reason for approval, then renders only the approved cuts via ffmpeg. Requires a word-level transcript.

Membangun motion graphic sinematik di browser (promo video, opener, intro, bumper, kinetic typography, explainer/video penjelasan bergambar 16:9 atau 9:16) memakai HTML + CSS + GSAP (+ Three.js bila perlu), dengan hasil yang bergerak seperti video sungguhan — bukan slide presentasi. Gunakan skill ini setiap kali user meminta "video promo", "opener", "intro animasi", "motion graphic", "bumper", "kinetic typography", "animasi teks sinematik", "explainer", "video penjelasan" (jurnalisme visual, kartun edukasi, kolase), atau menunjukkan referensi video promosi/explainer dan ingin dibuatkan versi web-nya — bahkan bila mereka tidak menyebut kata "motion graphic" secara eksplisit. Juga gunakan bila user mengeluh hasil animasi web "seperti PPT/presentasi" dan ingin lebih sinematik, atau ingin merender animasi web menjadi file video MP4, atau ingin explainer kartun/animasi dibangun LANGSUNG di After Effects lewat bridge/MCP Higgsfield.

Создаёт видео из HTML-анимаций. Генерирует несколько вариантов дизайна в HTML, рендерит каждый в видео (покадровый захват + ffmpeg), анализирует кадры, выбирает лучший вариант и собирает финал, перенимая удачные решения из остальных. Используй, когда просят сделать ролик, промо, анимированный баннер, заставку, презентацию-видео или «видео из HTML».

用户个人的视频剪辑完整流程与资产包，用于把一批视频素材（旅行/vlog/运动相机/手机/网盘大文件等）剪成一条成片。当用户要求"剪辑/做个视频/做 vlog/剪成一条/成片/回忆视频"，或素材很多很大、需要由助手侧完成下载与全部后期、只交付成片时使用。包含标准流程、可复用脚本（下载守护、系统恢复、抽帧网格、调色剪辑、拼接、终验）以及完整复盘与问题排查手册。具体滤镜/色调不固定，按用户每次要求现场确定。

Export social video variants, inspect playable media, review thumbnails and caption placement, and prepare publishing assets.