Explore

Converges an underspecified request into a typed IntentSpec before planning or acting. Use when the user asks to implement, add, change, refactor, fix, migrate, configure, handle, triage, sort out, look into or decide something and the request leaves decisions open (which objects, which approach, what happens on failure, which trade-off), contradicts itself, admits two readings, or rests on an approach its sources may rule out — in a codebase, a ticket queue, a research brief or a runbook — or when the user says "clarify the intent", "what do you need from me", or invokes intent-router. Looks up what its sources hold (code, history, decision records, a ticket log, an order record, the policy in force) before asking; asks only preference or irreversible questions, one at a time, with a recommended default; halts instead of guessing; checks the delivered work against the spec. Silent on a fully stated task its sources do not contradict. Not for explaining existing state ("what does X do", "why is Y slow").

Make an explainer video (MP4) with a voice-over using explainroo. The voice is made on this computer and no API key is needed. Use when the user asks for an explainer, how-to, walkthrough, launch or tutorial video, an animated diagram with a voice-over, or a short for YouTube, Reels or TikTok.

Make 100 versions of Claude fight to the death over one task. Spins up N sub-agents (default 100, --quick for 16), gives every one the exact same task plus a different strategy card (reasoning mode, workflow, strategy), then runs a single-elimination bracket: they attack each other's solutions, defend and revise, and a judge scores every match on a written rubric until one solution survives. Use when the user is not satisfied with an answer, calls it a bad answer, says try again or do better, says "arena", or asks to make them compete.

Finds the price buyers will actually pay. Reads the consumer panel's four price answers per buyer into a Van Westendorp price-sensitivity range (too cheap, a bargain, getting expensive, too expensive), sets it against what competitors really charge and against the CFO's margin, and recommends a price, a price ladder (good / better / best) and an opening offer to test. Use when the user says "how much should I charge", "price this", "is $X too expensive", "pricing strategy", "what price", or after the consumer panel.

Your business planner. Compiles everything the other founder skills wrote (the board's verdict, the competitors, the buyer panel, pricing, the offer, the CFO's numbers, marketing, brand, operations and the launch plan) into one business plan with a verdict at the top that is computed from the numbers, never written by hand: Profitable, Not yet (with the failing numbers) or Incomplete. Adds an executive summary and a one-page version for a partner, lender or investor. Use when the user says "write my business plan", "put it all together", "is it profitable", "give me the plan", or at the end of the founder pack.

Your operations manager. Works out how the business actually runs on day one: suppliers and what they cost, who works when, the step-by-step routines for opening, serving or shipping and closing, the tools to run it, the licences and permits to check, and a risk register with what to do when each thing goes wrong. Feeds real costs back to the CFO. Use when the user says "how do I run this", "operations", "suppliers", "staffing", "SOPs", "what do I need to open", "permits", or "day one".

Builds an offer too good to refuse. Lists every problem the buyer hits on the way to the result they want, turns each into a solution, trims to what is cheap to deliver and valuable to the buyer, and stacks it with honest bonuses, a guarantee the CFO can afford, real urgency and a clear name. Scores it on the value equation (outcome, likelihood, time, effort) and re-tests it on a quick buyer panel. Use when the user says "make my offer better", "why won't people buy", "what should I include", "guarantee", "bonuses", "grand slam offer", or after the panel shows people passing.

Your marketing director. Turns the buyer panel and the competitor map into positioning, the channels the target customer actually uses, a 30-day launch campaign with dated posts and ads, the first ten hooks, a budget split and the numbers to watch. Every message answers an objection real or simulated buyers raised. Use when the user says "how do I market this", "marketing plan", "launch campaign", "how do I get customers", "where do I advertise", "write my ads", or "positioning".

Your launch lead. Builds a dated, step-by-step launch plan: a cheap real-world test first (a waitlist, a pre-sale or a pop-up with a real price, so real buyers confirm what the simulated panel said), then the countdown to opening day with owners and deadlines, the launch-day run sheet, and the first 30 days of what to measure and what to change. Use when the user says "how do I launch", "launch plan", "when can I open", "launch checklist", "go live", or "soft open".

The consumer panel. Spins up a swarm of buyer sub-agents (default 100, --quick for 20) trained on the user's target customer, each with a different income, age, buying behaviour and objection, shows every one the same pitch, and runs the business through 100 buyer scenarios to see who buys, who doesn't, and why. Tallies the answers into buy rates by segment, the reasons people pass in their own words, what would flip a no, and price answers for founder-pricing. Use when the user says "would people buy this", "test it on customers", "run the consumer panel", "simulate buyers", "who is my customer", or "validate demand".

Maps every competitor and the gap they leave. Finds the direct competitors, indirect ones and the substitutes a buyer would use instead, from public sources only, and records what each sells, at what price, how they position themselves and what their customers complain about in public reviews, every fact with a link. Ends with a positioning map and the gap the new business can own. Use when the user says "who are my competitors", "is this market crowded", "what else is out there", "competitor analysis", "find the gap", or "who am I up against".

Your CFO. Builds the unit economics of a business before money is spent: what one unit earns after its direct costs, the real profit margin at the planned volume, the break-even point in units a day, year 1 month by month, the cash needed before it pays for itself, how long the startup spend takes to earn back, and what happens if the price drops, sales come in low or costs rise. Every number traces to an input the user can check. Use when the user says "find the profit margins", "do the numbers", "is this profitable", "break-even", "unit economics", "how much do I need to start", or "run the CFO".

Your brand designer. Names the business (with the trademark, domain and handle checks to run before falling in love with a name), writes the voice, a one-line promise and a tagline, and briefs the look: colours, type, logo and the first things a customer sees, so they match the positioning the marketing director chose. Use when the user says "name my business", "brand this", "logo", "brand identity", "what should it look like", "tagline", or "brand voice".

Your board of directors. Puts a business idea in front of a three-member board, each a sub-agent reading the idea through one published framework: the Offers lens (from Alex Hormozi's $100M Offers), the Monopoly lens (from Peter Thiel's Zero to One) and the Product lens (from Walter Isaacson's biography of Steve Jobs). Each member scores the idea, names the risks that would kill it and votes fund, fund with conditions, or pass. Use when the user says "test my business idea", "would this work", "pressure-test this", "what would a board say", "is this a good business", or starts the founder pack on a new idea.

各账号视频流水线共用的四道工序（事实核查与来源台账、AI 封面比例验收、平台文案机制、成片机器验收与交付边界、合规自查）。不直接触发；由各账号 skill（如本仓库的 investigation-video-skill）在对应环节引用。音色、字幕样式、品牌、时长、口吻、平台集合都不在这里，留在各账号 skill。

Diagrams and checkable docs from a codebase. Draws ERDs, C4 context/container/component views, API endpoint catalogues, lifecycle/state machines, and service-to-service call maps as a validated JSON spec plus a standalone HTML viewer, with Mermaid export. Imports OpenAPI, GraphQL SDL and Prisma; otherwise reads the code (NestJS, TypeORM, Express) and writes the spec. Use for: database diagram, ERD, data model, table relationships, C4 or architecture diagram, system context, API map, endpoint list, "show me the endpoints", status flow, state machine, allowed transitions, "what happens after X is approved", how services call each other, microservice map, Mermaid diagram. Also writes documentation as a fact graph: each statement carries its source and a computed confidence, and a drift check fails CI when the code moves. Use to document a system, write architecture or onboarding docs, or explain how a system works.

用 catbus CLI 读取或操作小红书（RedNote）、抖音、TikTok、B 站（Bilibili）、快手、微博、闲鱼、淘宝、京东、X（Twitter）这 10 个平台：搜索、详情、评论、用户、推荐流、下载、发布、点赞关注、私信、直播弹幕，输出统一的 JSON。用户要查询、采集、导出、监听或操作这些平台上的内容和账号时使用。

中文技术文档写作规范。写或修改中文技术文档、开发文档（README、设计文档、接口说明、教程）时使用。

Пишет unit-тесты для указанного файла или функции в стиле существующих тестов проекта. Используй, когда просят добавить тесты, покрыть код тестами или проверить граничные случаи.

Безопасный рефакторинг маленькими шагами без изменения поведения. Используй, когда просят отрефакторить, упростить, разделить модуль, убрать дублирование или почистить код.

Готовит заголовок и описание pull request по изменениям в ветке. Используй, когда просят описать PR, подготовить pull request или написать описание изменений.

Полный цикл исправления GitHub issue по номеру - анализ, тест, исправление, коммит.

Создаёт git-коммит в формате Conventional Commits по текущим изменениям. Используй, когда просят закоммитить или сделать коммит.

Ревью изменений с фокусом на безопасность по стандартам команды. Используй перед мерджем или по просьбе проверить безопасность.

Make a short Vox-style documentary explainer (30-60 seconds) entirely in code with Claude Code. Torn paper, halftone texture, bold condensed type, motion on twos. Use when someone says "make a Vox-style video", "explainer video about...", "why does X happen", or wants a documentary-style motion graphic for LinkedIn, Instagram or YouTube without After Effects or a video generator.

Make a cinematic 3D title sequence (8-15 seconds) in code with Claude Code and Three.js. A camera moves through a lit scene, such as a rainy street at night with brand names as neon signs, then a big question or title lands. Use when someone wants a scroll-stopping opener, a film-style intro, "3D intro", "cinematic opening", "title sequence" or a hook for a video, reel or newsletter.

Turn references you admire into a scene-by-scene Claude Motion prompt for your own topic and brand. Takes a video for the moves (screenshots, frames or a contact sheet of a reel or ad) and, optionally, a look reference for the style (a website, a brand, a moodboard). Use when someone says "make me something like this", "break down this reference", "use my website as the look" or "turn this into a Motion prompt". Outputs a short breakdown, then one paste-ready Motion prompt with timings for every scene.

Turn any horizontal or square motion video into a 1080 x 1920 Instagram Reel or TikTok that uploads cleanly. Headline and logos inside the safe zone, H.264, yuv420p, limited colour range, bt709, a silent audio track when there is no sound, and an ffprobe check before upload. Use when someone says "make this a reel", "vertical version", "export for Instagram", "TikTok version", "9:16 cut", "reel export" or has a 16:9 or 4:5 video that needs to go on a vertical feed.

Turn a newsletter edition into a 12-20 second promo clip for LinkedIn, Instagram or X that drives people to read and subscribe. Hook question, three proof frames from the edition, then the newsletter name and where to read it. Use when someone says "promo for my newsletter", "trailer for this edition", "get people to my Substack", "teaser video" or "promote this issue".

Write a complete 8-part Claude Motion prompt (The bar, Format, The one rule, Look, Motion, Never, Before you finish, The scene) for one short looping motion graphic, in your own brand. Takes your brand (a website screenshot, a logo, hex codes, or nothing for the Apple-style default) and one moment you want to show, in plain words. Use when someone says "write me a Motion prompt", "turn this idea into a motion graphic", "fill in the 8 parts" or "make my motion brief". Outputs parts 1 to 7 as a reusable brief and part 8 as the scene, in one paste-ready code block.

Build any of 16 premium motion effects from scratch in your own brand, as one 8-second looping HTML file with window.seek(seconds) so it exports frame-exact to MP4 or GIF. Interfaces (button to player, search to results, card to workspace, tabs to panels), data (chart morph, dashboard zoom, spring stack, magnetic dock), type (masked type, elastic type, text to layout, image reveal) and systems (perspective shift, glass focus, flowing paths, particle logo). Use when someone says "build the chart morph", "make a button that turns into a video player", "animate my logo with particles", "motion effect", "UI animation", "product animation", "animated dashboard", "kinetic type" or names any of the 16 effects.

Turns a rough idea for an animated graphic into a precise build brief for Claude Code, in your own brand. Asks for anything essential first (exact numbers, size, length, brand), keeps real results apart from sample data, and never invents a figure. Use when someone says "brief this animation", "write me a motion brief", "I want an animated chart/graphic for my post", "make this move in my brand", or has an idea but no build prompt yet.

Give the same motion brief to three AI models, keep each first try, render all three the same way and stack them into one comparison video with model names and logos. Makes a 16:9 side-by-side for a newsletter and a 1080 x 1920 stacked reel for Instagram and TikTok. Use when someone says "compare three models", "same prompt three AIs", "model showdown", "which AI is the best motion designer", "side by side comparison video" or wants a fair head-to-head of AI outputs.

Make a 12-second looping milestone clip in code with Claude Code. Thousands of tiny points drift like a night sky, pull together into your real number (followers, subscribers, revenue, customers), a line draws underneath and the label settles in. Premium, not a party. Use when someone says "milestone video", "celebrate 10k followers", "animate my subscriber count", "number reveal", "particle number animation" or wants to mark a result on LinkedIn or Instagram.

Turn a newsletter or blog cover into a seamless looping GIF at 896 x 640, 180 frames at 20fps. Only the one element you name moves, the loop is rotated so frame 1 shows the finished picture, and the seam is measured, not eyeballed. Use when someone says "animated cover", "make my cover move", "looping GIF cover", "newsletter cover GIF", "animate my thumbnail" or wants a cover that moves in the inbox without looking busy.

Make a 30-45 second product launch video in code with Claude Code, the kind SaaS companies post on launch day. One sentence split across scenes, real product screens with slow camera moves, a typed prompt, a proof beat and a clear call to action. Use when someone says "launch video", "promo video for my product", "announce my course/cohort/offer", "make a launch reel" or "trailer for my product".

The setup step every other skill in this pack reads. A short interview plus 3 to 5 reference frames becomes two files in your project, brand.md (who you are, what you sell, your assets) and MOTION.md (your colours, type, timing and motion rules), plus a rule in CLAUDE.md so Claude reads them before it animates anything. Use at the start of any motion project, or when someone says "set up my brand", "brand intake", "build my motion design system", "make it look like my brand", "onboard me", "write my MOTION.md" or drops a batch of brand frames into chat.

Rebuild an Apple-style Mac launch film entirely in code with Claude Code. A thin menu bar, the black notch growing into widgets one scene at a time, full-screen gradient wallpapers and one short caption per scene. No screen recording, no After Effects. Use when someone says "Apple-style launch", "Mac launch video", "notch animation", "make it look like an Apple keynote", "macOS UI launch film" or wants a product launch that looks like native Mac interface.

Make a looping animated chart for LinkedIn or a newsletter in code with Claude Code. Bars that become a line, a number that counts up, a before and after. Every value stays exactly as given. Use when someone says "animate my chart", "animated graph", "make my results move", "stat animation" or wants a data graphic that loops.

Create or improve procedural risograph still art, illustrations, posters, print series and images in this repo. Develops reference-led composition, perspective, pose and ink plates; delivers a native-resolution PNG and self-contained Canvas 2D source. Use for standalone still requests; use riso-film for moving work.

Scores a riso film or reworks its score. Designs the cue list from the film's FILM.md events, builds a deterministic renderAudio() from the sound kit in studies/sound.html, iterates with tools/audio.mjs, gates determinism, length, loudness and sync, and records measured results in FILM.md. Use for "score this film", "add sound/music", "the climax is too loud", "fix the handoff at 16 s", "check the mix", sound effects, foley or a soundtrack on a riso film. Not for drawing or motion (riso-film) or changing tools/.

Design, build or improve procedural risograph animations and videos in this repo, including narrative films, abstract shorts, motion effects and visual inspection. Delivers self-contained Canvas 2D HTML and a verified MP4. For standalone still art use riso-still; for sound use riso-score. Not for editing tools/.

Run Jev Review as a repeated scalar feedback loop during nontrivial coding work. Establish a score baseline after a coherent implementation, diagnose weak dimensions yourself, improve the code, validate it, and rescore with the previous evaluation until important metrics improve or no further justified change remains.

Make a promo / motion video for a software product with HyperFrames, using the saas-motion-kit 7-stage process (brief → components → theme → storyboard → build → sound → deliver), the clean-or-imaginary component rule, the creative muscle (tone matrix, variety rules, component forge, transition atlas, variety audit) and the 101-theme gallery in docs/. Use when someone asks for a launch video, product promo, booth loop, feature reveal or social motion piece for an app, SaaS or developer tool.

Make videos as web pages with OpenFilm. Use when the user asks for a video, motion graphics, an animation, a launch film, an explainer, a music video, to edit, cut, check or render a film (film.html and pages with window.film), or to open OpenFilm or its Studio.

Read or change compact-adviser settings for Codex - the mode (hint or off), the minimum context before a checkpoint is judged, the context budget, TypeSafe request logging, and the TypeSafe API key. Use when the user asks about compaction advice, the /compact hint, compact-adviser status, or wants to turn the adviser on or off.

Use when changing or evaluating judge prompts, scoring thresholds, usage floors, or judgment eval corpora.

Use when modifying or testing `packages/codex-plugin` or Codex `/compact` behavior.

为产品、网站、App、SaaS、开源项目、课程、AI 模型或活动制作有商业质感的宣传片（4K MP4 + 网页版），流程包括调研、需求问询、风格样帧、真实页面 4K 截图、配乐卡点、AI 旁白、混音和自动校验。用户说「做宣传片」「做产品介绍视频」「做推广视频 / 预告片」「给这个项目做个视频」，或者要求 promo video、product trailer、launch video 时使用。

Build new interface components and bounded composition patterns inside an existing project. Use for named components, interface blocks, and workspaces. Preserve the target project's design system and stack; for a complete new page with page-level structure and multiple sections, use Pagina page-builder instead. Do not use to audit or reconstruct an existing component, plan a whole product or website, or handle purely editorial copy requests.

Use Machinist to create, assign, and monitor software tasks. Use when a coding agent needs to work with Machinist, its GitHub issue workflow, intake labels, direct runs, or managed queue.

Report in one line what the Claude Code, Codex and Mistral Vibe sessions on this computer are doing right now (how many are working, how many need the user, whether it is the user's turn). Use when the user asks what their agents or sessions are doing, or whether anything is waiting for them.

Open Agent Standup, a local page that shows the Claude Code, Codex and Mistral Vibe sessions on this computer as a live team call (who is working, who waits on whom, who needs the user). Use when the user asks to watch, see or follow their agents, sub-agents or sessions, or asks for the standup or the call.

Use when code needs a judgment about natural language that rules or regex cannot make (classify, route, triage, moderate, score, rank, match, dedupe, filter, extract, verify, or gate an action), when brainstorming where AI could fit in an app, when an LLM call exists only to pick a label, score, or yes/no, or when code uses TypeSafe AI, Jev, System One, typesafe-sdk, or @typesafe-ai/sdk.

Generate or edit images with Codex's built-in image generation from any agent, through an MCP tool or a plain CLI. For every image request it first proposes at least four divergent style directions (24 distilled templates, 48 presets, 20 Mondo poster artists), lets the user pick, expands the prompt, generates, and checks the result. Scenario craft for 视频封面, 视频海报, 小红书配图, 公众号封面, X 封面, 朋友圈海报, 书籍封面, 专辑封面, 产品海报, 人物写真, 节气海报, 展览海报, PPT 页面, 图生图. Use for 生图, 配图, 海报, 封面, "给我几个风格", "用 Codex 画一张". Excludes: editing photos with local tools, charts or diagrams from data, UI mockups, video generation, or anything that needs an API key-based image service.

ショーリール品質（Behance / OFFF / Awwwards級）のモーショングラフィックス動画を HyperFrames（HTML + GSAP + SVG + Canvas、Apache-2.0）で制作する。remotion-motion-reel の HyperFrames 版で、Remotion の Company License が不要な構成。テーマ・素材（キャラクター画像、音声など）から世界観・コピー・デザインシステム・タイムライン・サウンドを補完し、120BPM/60fps で拍同期した映像を実装・レンダリングまで行う。「HyperFramesでかっこいい動画」「ライセンスフリーでモーショングラフィックス」「ショーリールをHTMLで」などの依頼時に使用。

Operate a simfleet device fleet: slimmed iOS simulators and Android emulators, parallel React Native worktree lanes with leased Metro ports, the native build cache, and per-agent device claims. Use this skill whenever a project has `.sim-fleet/project.json` and an agent needs to boot or slim a simulator or emulator, run one or more worktrees on devices, allocate or inspect Metro ports, launch an environment/build variant, tap/type/inspect a device, claim a device for its session, or diagnose simulator RAM — even when the request only says "run this branch", "open the simulator", or "fix a Metro port collision".

Design or change the person's Claude Code transcript skin with the skins mod. Use when they ask to restyle, recolour or theme Claude Code, make a skin from a mood, a brand, a wallpaper or another theme, change the spinner or footer words, or switch the rail, tables or shimmer on or off.

Use when the user asks to pick a model, subscription, or reasoning effort, or to run router status, usage refresh, or resume a router session. Also use when you were launched by model-router (your task has a "Router session:" line) and your phase is complete, to route the next phase. Invoke the model-router CLI instead of choosing a model yourself.

Build AI-powered software with TypeSafe: small units of AI intelligence you can use like programming primitives. Its System One models, including Jev, turn natural language and application state into typed judgments and probabilities that code can combine. Use when a feature needs programmable common sense, when brainstorming what AI could make possible in an app, or when an LLM prompt-and-parse step could become a structured decision. Applications include routing, ranking, extraction, verification, and interactive experiences; these are starting points, not the limits. Read live docs and cookbooks to find useful patterns and discover new combinations.