Video & Animation

Maskotlu (Demli) 15-20 dakikalık Türkçe sohbet videosu üretir. Opus yalnız plan.md yazar, render ve kontrolü Sonnet alt ajanı yapar. "Demli bölümü", "perde ile video", "uzun maskotlu video", "yeni bölüm yaz" dendiğinde kullan.

Turn an idea or script into a finished animated video (Bible story, tech/programming, maths/education or general story/ad) in a chosen style (stickman, whiteboard, flat 2D, modern vector, storybook, toon 3D, cinematic 3D) and format (9:16, 16:9, 1:1, 4:5), after checking that the computer can render it safely. Use for any request to make an animation, animated explainer, cartoon, Bible video or animated short.

Run Claude as the technical project lead of a video game (or any creative project) for a director who does not code. Covers communication, agent and model management, token economy, testing, tools to set up, visual and audio harmony, UI and icon rules, avoiding the "AI look", balancing, playtesting and reporting. Use at the start of every work session on such a project.

Direct and produce finished videos using AI artwork with editable SVG foregrounds, real product UI, or procedural 3D. Use for explainers, product demos, launch films, Shorts and before/after studies needing readable visuals, timed annotations, smooth capture and a local review MP4. Not for an isolated image or ordinary website animation without a video deliverable.

Edits a raw talking-head video (Reels / TikTok / Shorts) into an engaging short-form edit from inside Claude Code. It removes pauses and bad takes, then adds motion graphics, infographic cutaways, zoom punches, emoji bursts and sound effects, and renders an MP4 ready for CapCut. Built Mongolian-first (script-to-audio alignment) but works in any language. Use this skill whenever the user wants to cut, edit or "make engaging" a talking-head, selfie or vlog clip, add motion graphics / infographics / B-roll to a video, or types /goat-video-editor, /video-edit, /видео-эдит, /эдит, or says "видеог минь эдитлээд өг", "энэ бичлэгийг засаад өг", "motion graphic нэм", "edit my video", "cut my talking head", even if they don't name the skill.

Turn any story idea into a finished narrated cartoon Short (9:16, 1080x1920, 30 fps, usually about 60 s) with original characters, places, camera moves, explainer graphics, captions, voices, music and sound effects, rendered locally in code (three.js in headless Chrome) with no AI video credits. Use when someone wants to make an animated story short, cartoon short, explainer, fable, history oddity, true-event story or what-if video from an idea; or wants to change a film made with this pipeline (characters, shots, camera moves, sound, captions, framing).

Make narrated faceless YouTube Shorts / TikToks / Reels fully in code on the user's computer, with a free local AI voice and word-synced captions, $0. Six styles - Deep Dive 3D (Zack D. Films-like 3D "what happens if" body/science explainers), Night File 3D (unsolved mysteries, true crime), Inside Out (flat 2D x-ray explainers), Stick Story (stick-figure "what would you do if" survival steps), Paper Atlas (history in cut paper with maps), Tall Tale 3D (funny true stories with flat 2D cartoon people in a sunny 3D world, one-word captions). Use whenever someone wants a faceless short, a what-if or what-happens-if explainer, a Zack D. Films style video, a mystery or true-crime short, a stick-figure story, a history short, a funny true story or storytime cartoon short, or a video in one of these styles, even from just a topic. Can also upload the finished short to the user's YouTube channel through Zapier when they ask.

Planifica y valida cortometrajes cinematográficos 3D: narrativa, storyboard, dirección artística, plan técnico, voz, subtítulos y control de calidad. Úsala al preparar un paquete de producción; no autoriza por sí sola generar contenido audiovisual.

Turn an article, URL, pasted text, or a specified section into 1–3 short silent explanatory GIFs with editable scene source. Use when the user wants animated article illustrations or text-to-GIF visual explanations; not for a static diagram or narrated video.

Turns a motion video that was made with code (HTML canvas, JS/Node or Python renders, SVG, any frame-by-frame renderer) into a native, fully editable After Effects project (.aep). The project has live text layers, shape layers, one precomp per object, a camera null with real keyframes, effects as toggleable layers, audio stems and narration markers. It's the AE equivalent of a layered PSD. Use this whenever the user wants an After Effects file, an .aep, "an editable version", "layers I can change", a project to hand to a motion designer or editor, or wants to tweak, retime or reuse a video that was rendered from code. Use it even if they only say "AE", "After Effects project", "make it editable" or "like a PSD but for video".

Creates an original faceless vertical reel or Short (9:16, 1 to 2 minutes) from start to finish: analysis of a reference reel, script (hook, explanation, call to action with a guide and a keyword), cloned ElevenLabs voice that is never cut, optional Atlas Cloud opening shot, one-shot Remotion animation with a mascot and word-by-word captions, a click-worthy cover set as the first frame, and an ending that always makes the same three moves: subscribe, link in bio, keyword. Use this skill whenever the user wants to create, remake or take inspiration from a faceless reel, Short or TikTok (\"make a reel like this one\", \"here is an Instagram link, make me the same\", \"a reel about the 5 tools...\", \"turn this topic into a reel\"), even without saying \"faceless\", and also to rework or fix a reel already made with this skill, its cover or its keyword.

NAVI (natural AI demo video) makes natural-sounding, faceless demo videos, walkthroughs, product tours, tutorials, screencasts and course lessons of any website or web app, with an AI voiceover that doesn't sound robotic. Claude explores the site, writes a talking script and clean captions, voices it with ElevenLabs, records the browser clicking in time with the voice, blurs private data, and exports an MP4 with captions. Use whenever someone wants a demo or explainer video, a screen recording with voiceover, a narrated walkthrough or a faceless tutorial, even if they don't name this skill. Needs Claude Code running on the user's own computer.

Clean up a raw screen/OBS/talk recording into a tight professional upload - transcript-driven cuts, dead-air and visual-static removal, off-limits personal moments, on-screen leak redaction, audio enhancement, strong opening, verified render, then chapters/title/tags/description/subtitles/thumbnail, delivered to ~/Downloads and opened. Use when the user says "edit the video I recorded", "cut the silence out", "remove embarrassing parts", "make sure nothing leaks", "make it YouTube ready". Not for building promo/motion-graphics videos from slides; thumbnails alone go to youtube-pack.

Make a vertical 9:16 reel in the fast-cut style - real full-screen video clips cut on the spoken words every 1-2 s, big condensed words and cards on black, a running number that drops as the story goes, humour from timing (reaction clips, freeze frames, punch-ins, cuts to silence), a dry funny voice, about 40-60 s. Use when the user asks for a fast-cut, Fireship-like or meme-style reel, a reel with real or funny clips, a reel like the Egyptian pound reel, or to fix such a reel.

Write, rewrite or review marketing and promotional copy with tone-of-voice control: landing pages, product pages, ads (Meta, Google), emails and subject lines, social and TikTok captions, headlines, taglines, CTAs, video scripts, and adapting copy between French, English and Hebrew. Use whenever the user wants persuasive copy, pastes a copywriter-settings block (from ailive.fr/copywriter), names tone settings ("hype 90", "more premium", "plus naturel"), or asks to punch up, humanize or check cla

Review existing marketing copy against ten copywriting laws and score it 0-100, with a word-for-word quote, a severity and a rewrite for every flaw. Works on web pages, landing pages, ads, video scripts and email, in English or Swedish. Use when asked to review copy, audit copy, critique a headline, check a landing page, explain why a page does not convert, or assess a client page before it is rewritten. Does not write new copy from scratch.

Build single-file HTML proposals, pitches and explainers in the GOV.UK Design System visual style, made of visual components rather than long prose. Use when someone asks for a proposal, pitch, one-pager, explainer or leadership-facing page "in the GOV.UK style", or wants a self-contained HTML document to share and comment on.

Apple's industrial design language as concrete rules for hardware and robotics CAD — form, continuous-curvature corners, edges, seams and gaps, CMF, vents, controls, status lights, docks, mechanisms, robot motion, and Apple-style renders. Use when the goal is aesthetic or premium craft — "make it look like Apple designed it", Apple-style, Jony Ive, unibody, squircle / continuous corner / G2 fillet, premium enclosure, restyle or design-review a robot or product from a CAD screenshot, render, STEP

作为 script-to-storyboard-prompts 的下游预演适配技能，接收其有效正式母稿、段内预算、资产档案与首尾状态，转成可调整的 Three.js 三维分镜并用 HyperFrames 导出参考视频。适用于用户要求把已交付分镜做成 HTML/3D 预演；不重做上游七步创作，不凭“继续”自动生成媒体。

Curated 64-repository Three.js and 3D video production tech stack with battle-tested libraries and a Before/After reference guide. Use this skill whenever planning, building, rendering, reviewing, debugging, or advising on Three.js scenes, 3D explainer or documentary videos, WebCodecs MP4 export, camera math, glTF optimization, post-processing, ambient occlusion, lip-sync, 3D characters, audio loudness normalization, 2.5D parallax, geospatial maps, physics fracture, or Minecraft 3D films—even if

Tulis caption siap unggah untuk konten sosial media (carousel, reel/short/TikTok, foto, story, testimoni, promo, thumbnail) di banyak platform sekaligus (Instagram, TikTok, LinkedIn, X, Threads, Facebook, Pinterest, YouTube Shorts, WhatsApp Status) dengan hook, isi, CTA, hashtag, alt text, komentar pertama, dan alternatif hook, disesuaikan dengan suara brand. Skill ini memeriksa batas karakter tiap platform, panjang hook sebelum "selengkapnya", jumlah hashtag, link yang tidak bisa diklik di Inst

Ubah brand board, logo, foto, atau deskripsi brand menjadi satu profil brand baku (profile.json) yang menjadi sumber tunggal bagi semua skill konten lain (command gambar, video Flow, teks dan layout, jadwal, caption, balasan, QA). Mengekstrak nama, tagline, CTA, palet warna dengan role dan hex, font, deskripsi logo, maskot, layanan, paket, nada suara, dan hal yang dihindari; memeriksa kontras warna (WCAG) untuk menemukan pasangan warna yang tidak boleh dipakai sebagai teks; menghasilkan Brand Lo

Periksa kualitas gambar dan video hasil ChatGPT/Google Flow/AI lain terhadap brand dan platform, lalu tulis prompt perbaikan siap tempel. Mengecek ejaan teks pada gambar, kepatuhan warna terhadap palet brand, teks yang masuk zona tidak aman (tertutup UI story/reels/feed), ukuran dan kontras teks, rasio dan resolusi, logo atau karakter yang berubah dari acuan, anatomi janggal, serta pada video meliputi lonjakan atau glitch antar-frame, drift warna, teks yang berubah bentuk, audio, dan durasi. Gun

Diretor de criação da agência de performance. Cria e itera criativos de anúncio (ganchos, textos, imagens estáticas, carrosséis, vídeos curtos) para Meta, Google (RSA, PMax, YouTube) e outros canais, a partir do comprador real e da oferta; organiza a matriz de ganchos, os formatos e tamanhos por posicionamento, diagnostica por que um anúncio não performa e garante identidade visual sem "cara de IA". Use quando o pedido falar de criativo, anúncio, arte, vídeo para anúncio, Reels/Stories, carrosse