设计与三维建模

Act as the Orca orchestration coordinator that drives GitHub issues (or free-text ideas) from rough idea to a merged PR with supervised Orca workers running /design-gh-issue, /specify-gh-issue, /build-gh-issue and /code-review. Use when the user hands off issues or ideas to be designed, specified, built and reviewed, says "orchestrate" or "hand off and supervise", or wants several issues worked in parallel.

/ow-start-transcribe: transcribe a recorded call on this machine, with speakers separated and confirmed by you (ow-tools, Claude Code only). Use for 'use ow-tools to transcribe intake/<client>/audio/', 'transcribe the fit call', 'transcribe the recording', or a workshop or demo recording. The recording never leaves the machine. NOT for writing the recap or intake record (that is ow-start's extract step).

The studio's Packaging Lead. Owns stage 6 for the studio-director, turning the finished video into everything a viewer sees before and around it, which covers the title and thumbnail as one pair, captions/subtitles, description, chapters, tags, hashtags, pinned comment and end screen. Delivers a Package Handoff. Picks and runs yt-package, arabic-titles-thumbnails, youtube-title-creator, arabic-captions, captions, video-caption-creation, yt-seo, arabic-youtube-seo, yt-chapters and nano-banana-image-generator. Use when the director reaches packaging, or when the user asks for a title, thumbnail, captions, description, chapters, tags or "everything I need to upload", or says "عنوان وثمبنيل", "اكتبلي الوصف", "شابترز", "كابشن للفيديو", "جهّزلي الرفع".

The studio's Packaging Lead. Owns stage 6 for the studio-director, turning the finished video into everything a viewer sees before and around it, which covers the title and thumbnail as one pair, captions/subtitles, description, chapters, tags, hashtags, pinned comment and end screen. Delivers a Package Handoff. Picks and runs yt-package, arabic-titles-thumbnails, youtube-title-creator, arabic-captions, captions, video-caption-creation, yt-seo, arabic-youtube-seo, yt-chapters and nano-banana-image-generator. Use when the director reaches packaging, or when the user asks for a title, thumbnail, captions, description, chapters, tags or "everything I need to upload", or says "عنوان وثمبنيل", "اكتبلي الوصف", "شابترز", "كابشن للفيديو", "جهّزلي الرفع".

Design low/med/high infrastructure with mandatory sources, diagrams and preserved decisions. /palarran:infra:plan; --plan uses no SSH or target changes.

Design and implement task-oriented MCP tools with clear schemas, compact complete results, permissions and independent evals. Use for /palarran:mcp:build; --plan performs no execution, auto install/activation or paid calls.

Create responsive visual design from UX concepts and references, choosing among ten established design systems or custom references. Use for /palarran:ui:design; creates visual previews, tokens and components, generates required images and preserves decision explanations for later tasks.

Design responsive application UX from a business description or README with research into 3–4 analogs, diagrams for every screen and states. Use for /palarran:ux:design; preserves concepts and decision rationale for later tasks. Does not define visual style or implement the interface without a separate request.

Generate patent-office-compliant figures via the PatentFig AI API — patent line art from text (PNG or SVG), vectorize drawings to SVG/DXF/vector PDF, AI-upscale images, and convert to filing-ready TIFF/PDF/PNG. Use when the user asks for patent drawings, patent figures, invention diagrams, USPTO/EPO-compliant line art, vectorizing a drawing for CAD, or preparing figures for a patent filing. Requires a PATENTFIG_API_KEY environment variable.

PocketBase 的集合与数据结构设计。用于创建集合、设计 schema、添加字段、设置关系，或在 base/auth/view 集合类型中选择。避免错误字段类型，记录零值默认行为，并涵盖关系级联。

PC 和主机游戏开发原则。引擎选择、平台特性、优化策略。

Motion-design video ads from a product URL or brief: script, storyboard, voice-over, music, MP4.

Create or improve a 32×32 pixel-art persona face for the Frontier Faces demo (examples/04-frontier-faces/personas.mjs) — the animated agent avatars rendered by face-term.mjs / the browser <cotal-face> / the gallery. Use when the user asks to add a new persona/face/avatar, improve an existing one (hair, eyes, likeness, expression), make it "look more like" someone, or fix a face that "doesn't read". Drives the example's own tools (img2rows.mjs, render-png.mjs) and follows FACE-DESIGN.md; uses a multi-agent variant loop for likeness.

Read and write Placard sidecar digests — `*.placard.md` markdowns beside expensive artifacts (decks, PDFs, spreadsheets, transcripts, image-heavy docs) that let an LLM understand a project cheaply. USE WHEN READING - a folder already contains `PROJECT.placard.md` or `*.placard.md` sidecars, or a session was told placards are present - read the map and the digests' inline `## Full text` instead of opening the originals; open an original only when a digest's "Open the original if…" contract fires or its `source_sha256` is stale. USE WHEN CREATING - the user wants to "digest" a folder, "summarize files for later", "make a project readable to an LLM", build or refresh a project map, or check whether digests are stale. Also use proactively before deep work in a folder full of decks/PDFs/spreadsheets.

Use when the user asks to plan a task or feature before implementing, create a workspace, write a design doc or ADR, break work into sessions and tasks for autopilot, resume an active workspace, or integrate workspace artifacts into docs.

Optional: produce the GSD design contract before plan-phase when the phase requires a spec.

Optional: produce the GSD UI design contract before plan-phase when the change touches the UI.

Complete browser automation with Playwright. Auto-detects dev servers, writes clean test scripts to /tmp. Test pages, fill forms, take screenshots, check responsive design, validate UX, test login flows, check links, automate any browser task. Use when user wants to test websites, automate browser interactions, validate web functionality, or perform any browser-based testing.

Find design references and video breakdowns for your next post. Free account: 100 credits a month.

以 HTML 或 Markdown 形式生成维护者级别的 PR 或贡献报告。适用于被要求深入审查 PR、解释设计、比较系统或准备合并建议时。

以 React（.jsx）组件形式创建达到 Awwwards 品质的高端网站设计，使其看起来像由每个项目收费 $50k+ 的顶级机构打造。每当用户要求网站、落地页、作品集或 Web 组件应看起来昂贵、高端、奢华、编辑风格、高档或具有机构品质时，使用此技能。当用户提到 Awwwards、FWA、获奖设计，或引用 Apple、Aesop、Bottega Veneta、Stripe 等品牌，或任何奢华/时尚/设计导向品牌时，也应触发。当用户说“让它看起来专业”“让它看起来昂贵”“我希望它看起来真的很好”“高质量设计”“不要千篇一律”，或对典型的 AI-generated 网站美学表示不满时触发。不要为仪表盘、管理面板、内部工具或美学次于实用性的功能性 UI 触发——标准前端设计技能会处理这些。

Prepares an Evolvix coach for a call or check-in review with one client: what changed since last time, the check-in answers that moved most, follow-through on training, food and habits, open questions, what the coach flagged and the last call's agreements, then three questions to ask and one suggested agreement. Starts the call when the coach begins, keeps notes and agreements while they talk, and finishes with a private note, tasks and a chat summary that stays a draft until the coach says yes. Also flags something to raise next time. Use whenever an Evolvix coach wants to prepare, start or finish a call, video call, check-in call or review with a client, asks what to cover with them, or says to bring something up next time, even if they never say brief. Dutch: "bereid mijn gesprek met Dennis voor", "check-in gesprek", "wat moet ik bespreken", "bespreek dit volgende keer", "gesprek afronden".

分析并设计定价策略，涵盖定价模型、竞品价格分析、支付意愿估算和价格弹性。适用于制定价格、评估定价模型、准备调价，或比较免费增值与付费模式。

TypeScript 后端的 Prisma ORM 模式：数据库结构设计、查询优化、事务、分页及关键陷阱，包括 updateMany 返回数量而非记录、$transaction 超时、migrate dev 重置数据库、批量写入跳过 @updatedAt，以及无服务器环境连接耗尽。适用于编写 Prisma schema 或查询，或调试事务、迁移及无服务器连接限制。

TypeScript 后端的 Prisma ORM 模式：数据库结构设计、查询优化、事务、分页及关键陷阱，包括 updateMany 返回数量而非记录、$transaction 超时、migrate dev 重置数据库、批量写入跳过 @updatedAt，以及无服务器环境连接耗尽。适用于编写 Prisma schema 或查询，或调试事务、迁移及无服务器连接限制。

Optimize LLM cost and latency with prompt caching: cacheable prefixes, cache-control headers, context layout, and cache-aware prompting. Use for high-volume apps.

Design effective prompts for LLMs: role framing, structured output, chain-of-thought, few-shot, and evaluation. Use to get reliable model behavior.

Local-first AI video editor: recordings to a finished film, cut by transcript, then verified

Build throwaway code that answers one design question that talk cannot settle. `logic` builds a self-contained HTML demo in docs/prototypes/ that drives a pure state module through free-play buttons and guided walkthroughs. `ui` builds 3–5 structurally different variants on an existing route behind `?variant=`, with a floating switcher, on a `prototype/<slug>` branch. `page` serves the demo, or a desktop and phone screenshot comparison of the variants, on the LAN. The answer goes into a short ADR (or into the proposed ADR of the `design` run that asked), the UI variants are deleted, and it offers `/delegate` for the proper build. Never commits. User-invoked only, through "/prototype".

Build a throwaway prototype to flesh out a design before committing to it. Routes between two branches — a runnable terminal app for state/business-logic questions, or several radically different UI variations toggleable from one route. Use when the user wants to prototype, sanity-check a data model or state machine, mock up a UI, explore design options, or says "prototype this", "let me play with it", "try a few designs".

构建一次性原型以回答设计问题。适用于用户想快速检验某个状态模型或逻辑是否合理，或探索UI应呈现何种样貌的场景。

用于处理 DICOM（医学数字成像和通信）文件的 Python 库。适用于读取、写入或修改 DICOM 医学影像数据，从 CT、MRI、X 光和超声等影像提取像素，匿名化 DICOM 文件，处理元数据和标签，将 DICOM 图像转换为其他格式，处理压缩 DICOM 数据或医学影像数据集。覆盖医学影像分析、PACS 系统、放射科流程和医疗影像应用。

多目标优化框架。NSGA-II、NSGA-III、MOEA/D、Pareto 前沿、约束处理、基准测试（ZDT、DTLZ），用于工程设计与优化问题。

面向医疗器械公司的ISO 13485质量管理体系专家。提供QMS实施、维护、流程优化和合规专业知识。适用于QMS设计、文档控制、管理评审、内部审计、纠正措施和ISO 13485认证活动。

构建或维护ISO 13485医疗器械QMS：设计、文档、CAPA、审计准备。每当用户面临13485认证审计或需要质量程序时使用。

Quarkus 3.x LTS 架构模式，涵盖使用 Camel 的消息传递、RESTful API 设计、CDI 服务、使用 Panache 的数据访问以及异步处理。适用于采用事件驱动架构的 Java Quarkus 后端工作。

Quarkus 3.x LTS架构模式，涵盖使用Camel实现消息传递、RESTful API设计、CDI服务、使用Panache进行数据访问以及异步处理。适用于采用事件驱动架构的Java Quarkus后端工作。适用于构建或审查Quarkus服务，尤其是涉及Camel消息传递或Panache数据访问时。

Quarkus 3.x LTS架构模式，涵盖使用Camel实现消息传递、RESTful API设计、CDI服务、使用Panache进行数据访问以及异步处理。适用于采用事件驱动架构的Java Quarkus后端工作。适用于构建或审查Quarkus服务，尤其是涉及Camel消息传递或Panache数据访问时。

优化 REA 提供方和共享分析工作流，以改善智能体上下文、延迟、CPU 和内存，同时保持正确性和证据。适用于分析瓶颈以及缓存或查询设计；CI/测试套件优化属于 rea-ci-tests-audit。

审计、调查、设计或变更 REA 调查工具、CLI/MCP 契约、提供方能力以及 Evidence 语义。

Deterministic Figma to Elementor, Gutenberg, Oxygen, Etch and HTML layouts for AI assistants.

Surgical code refactoring to improve maintainability without changing behavior. Covers extracting functions, renaming variables, breaking down god functions, improving type safety, eliminating code smells, and applying design patterns. Less drastic than repo-rebuilder; use for gradual improvements.

Run one named, scoped refinement pass on an interface whose design direction is already chosen: critique, polish, clarify, adapt, harden, bolder or quieter, touching only the named target and keeping the identity. Use when a person asks to critique, polish, tighten, harden, adapt, quieten or make bolder an existing page, screen or component.

Reframe a video to a different aspect ratio for social platforms — vertical 9:16 (Reels/Shorts/TikTok), square 1:1, or 4:5 portrait — from horizontal source. Use whenever the user wants to make a video vertical/square, fit it for Instagram/TikTok/YouTube Shorts, repurpose a 16:9 clip for social, or crop/pad to an aspect ratio. Covers crop-vs-pad choice, keeping an off-center subject in frame, and following a moving subject.

Search real website sections from hand-picked sites by look, for design references and inspiration.

使用 Remotion 从 Stitch 项目生成带平滑转场、缩放和文字叠加的演示视频

Remotion 的最佳实践与全面指南——在 React 中以编程方式创建视频，涵盖动画、合成和媒体处理

Remotion 的最佳实践：使用 React 制作视频。

Remotion のベストプラクティス - React で動画を作成する。3D、アニメーション、音声、字幕、チャート、トランジションなどをカバーするドメイン固有の29のルール。

Remotion 最佳实践 - 在 React 中创建视频。29 条领域特定规则，涵盖 3D、动画、音频、字幕、图表、过渡等。