Writing & Editing

Prepare Arabic (Egyptian dialect or MSA) voiceover and ad scripts so ElevenLabs and other TTS engines pronounce them correctly. Use whenever writing or fixing an Arabic script for text-to-speech, a reel/ad voiceover, an IVR prompt or an audiobook line; when Arabic TTS mispronounces numbers, prices, dates, percentages, brand names, English loanwords or domains; or when the user asks for Egyptian «عامية» narration.

Writes, checks, and edits documents in ASD-STE100 Simplified Technical English (STE, Issue 9). Use whenever the user mentions STE, ASD-STE100, Simplified Technical English, Simplified English, or controlled language, and also when they ask to write or rewrite procedures, work instructions, maintenance tasks, manuals, runbooks, user guides, or safety warnings so that the text is unambiguous, easy to translate, or easy for non-native English readers. Also use to audit or review an existing documen

Explain a bug, defect, incident, race condition, process breakdown, or confusing system behavior as a story told from the beginning, for a reader who does not know the code or the system. Use this whenever the user asks to be walked through an issue step by step, asks for it "like a story", says "start from the beginning", "explain it from scratch", "I'm not familiar with this", "I don't get what's happening", or asks how a failure happens after an investigation, even if they do not use the word

Load before writing, designing, approving or sending any customer-facing piece that carries hard facts: a post, story, reel, ad, flyer, caption, web page, proposal or chat reply that states a price, discount, phone or WhatsApp number, name, address, hours, date, shipping or payment terms, quantity, or a product claim. Use it for "check the prices", "is that the right number", "where did you get that", "¿de dónde sacas eso?", "ese no es el número", "ese precio no es", "no pongas mi nombre", "revi

Dünyadaki herhangi bir insan dilini öğrenmek için kullan: yaygın yaşayan diller, az konuşulan ve tehlike altındaki diller, klasik veya ölü diller (Latince, Osmanlıca, Sanskritçe vb.), yapay diller ve işaret dilleri. Kullanıcı bir dili öğrenmek istediğinde ya da o dilde öğrenme programı, ders, kelime listesi, dil bilgisi, telaffuz, yazı sistemi, alıştırma, seviye testi, konuşma veya yazma pratiği istediğinde, skill adını anmasa bile kullan. Hiçbir dile bağlı değildir: her dil için önce resmî kurum, dilbilim, sınav ve topluluk kaynaklarını internetten araştırıp analiz eder, sonra CEFR seviyelerine dayanan, kaynakçalı bir program hazırlar. Programlama dilleri için kullanma (o iş programlama-dili-ogret skill'inindir). Tek seferlik çeviri ya da tek bir kelimenin anlamı sorulduğunda da kullanma.

Proofreads and polishes academic writing in the humanities and social sciences while preserving the author's meaning, argument, voice, specialist terminology, and citations. Use when a user asks to check, correct, proofread, edit, or improve academic writing.

Crea el Documento Madre de Comunicación de una marca o de un lanzamiento, es decir, la tesis, el mensaje central que no rota, la frase de remate, los dos enemigos, los pilares y los ángulos, en español sencillo y una pregunta a la vez. Después escribe y revisa las piezas con él (reels, carruseles, historias, captions, anuncios, correos, WhatsApp, página, clase y objeciones de la llamada de venta) y arma la parrilla del mes. Úsala SIEMPRE que digan "/crear-documento-madre", "crea el documento madre de mi marca", "arma la línea de comunicación", "cuál es mi mensaje central", "mis enemigos y mis ángulos", "la comunicación del lanzamiento", "escríbelo con mi documento madre" o "revisa si este copy está alineado", o pidan un copy, guion, gancho, anuncio, correo o idea de contenido para una marca que ya tiene su Documento Madre guardado, aunque no nombren la skill. NO la uses para armar la oferta o los precios (crear-oferta), hacer la página web ni montar campañas en Meta Ads.

Rewrite an operating procedure, runbook, setup guide or safety instruction in a controlled-language style so a tired, stressed, non-native or first-time reader can follow it without misreading. Use it when someone asks to make steps clearer, write a runbook or SOP, prepare instructions for someone else to carry out, translate-proof a procedure, or "write it like a maintenance manual".

Use when shaping brand story and positioning. Ten lessons from Phil Knight's Shoe Dog — belief-first, community-first, product-as-ad, own the brand — each with an AI-marketing application.

Escreve copy de resposta direta em português do Brasil: landing pages, páginas de vendas e de captura, sites institucionais, e-mails e sequências de e-mail, prospecção fria (cold mail, LinkedIn, WhatsApp), roteiros de vídeo (anúncios, Reels, YouTube), VSL, artigos e advertoriais. Diagnostica público e mercado antes de escrever. Use sempre que o usuário pedir para escrever, reescrever, revisar ou melhorar um texto que precisa vender, gerar contatos ou levar a uma ação, mesmo que ele não diga "copy" (ex.: "faz o texto do meu site", "minha página não converte", "escreve uma sequência de e-mails de lançamento", "cria um cold mail pra prospectar clínicas", "roteiro de Reels pro meu anúncio", "script de VSL", "artigo pro blog que leve pro meu produto", "advertorial pra rodar no Meta", "write a sales email"). Não use para posts e legendas orgânicas de redes sociais nem para textos acadêmicos, técnicos ou literários.

Builds a tailored job application from a job posting and a resume - a rewritten resume and cover letter as Word files and PDFs that applicant tracking systems can read, plus a prep sheet PDF with a requirement-by-requirement fit check, keyword coverage, which portfolio pieces to lead with, outreach notes to the recruiter and hiring manager, interview stories, questions to ask and a follow-up schedule. Never invents experience. Use when someone asks to tailor a resume, write a cover letter, apply for a job, prepare an application, or reach out about a role. Usage - /application-kit followed by the job posting link or text, with a resume attached.

Builds a sponsorship or brand partnership pitch deck as a designed 16:9 PDF with speaker notes, from a brand and a property (venue, event, team, league, festival, tour, creator, platform or media property). Researches both sides, shows why they fit, proposes activation ideas, packages partnership tiers, and lays out how results will be measured. Use when someone asks for a sponsorship deck, partnership proposal, sponsor pitch, rights package, or help pitching a brand on a sponsorship. Usage - /sponsorship-pitch followed by the brand and the property, for example "Oatly x Coachella".

Writes ready-to-paste listings for selling or sharing a 3D print model file on marketplaces such as Etsy, Printables, MakerWorld, Cults3D and Thangs - titles, descriptions, tags, print settings, assembly notes, license, price based on comparable models, a photo shot list and a download README - and checks titles and tags against each marketplace's limits with a script. Flags trademark and character risks and AI-disclosure rules. Use when someone wants to list, sell, upload or price an STL or 3MF model, write a 3D model description, or set up a 3D print file shop. Usage - /print-listing-kit followed by the model name, with photos or files attached.

Audit a scholarly manuscript or bibliography for potentially hallucinated references, verify bibliographic identities against authoritative records, and produce a Markdown report with direct links for human checking. Use for citation checks, bibliography verification, and detecting fake, fabricated, or hallucinated references (参考文献チェック・文献の実在確認), including reference-existence and metadata verification. Not for judging whether cited sources support the manuscript's claims.

Review manuscripts drafted or edited with generative AI for recurring failure patterns - terms used before they are defined and prerequisites explained out of order; inconsistent data series, colors, and ordering across figures; overused parenthetical dashes and semicolons in English; unnecessary negation or contrast, including traces of the editing process such as leaked instructions or narration of what was changed; overly long paragraphs; excessive lists; bold or italic emphasis used for highlighting rather than for conventional purposes; and inconsistencies between newly edited or added parts and earlier parts of the manuscript. Use for pre-submission and post-editing quality checks of AI-assisted papers (AI生成・AI編集した論文原稿のチェック). This is a targeted check, not a full peer review, and does not assess novelty, methodology, or statistical validity. Report findings with evidence and precise locations, changing text, figures, or data only when asked.

Produces enterprise solution design documents, reference architectures and architecture decision records to formal industry standard. Use for solution design, infrastructure or application design, high level and detailed design, migration and remediation design, and reference architecture. Enforces a fixed numbered structure, a declarative specification voice, numbered and captioned tables and figures, evidence traceability, and explicit assumption, constraint, dependency, interface, decision and risk registers.

Rules for building application screens as work surfaces — purpose, every screen size, navigation, density, hierarchy, lists, forms, copy, states, ergonomics — plus a review procedure with a screenshot script. Use when building, changing or reviewing a screen, view, list, table, form, dialog or dashboard in a web or desktop app, or when a UI looks generic, template-made or AI-made.

Beginner-friendly, end-to-end playbook for programmatic SEO (pSEO): site inventory, fixing "crawled - currently not indexed" pages, keep/rework/redirect/delete decisions, pillar and content hierarchy, breadcrumbs, internal linking, page types and layouts, CMS requirements, developer and designer handoffs, content sheets, sheet-to-CMS publishing (Framer, Webflow, Strapi, WordPress), test batches, launch checks, indexing and lead tracking, with scripts for URL grouping, Search Console triage, near-duplicates, launch checks, link counts, sitemaps and sheet QA. Use it whenever someone wants to plan, build, clean up or launch pages at scale from templates or a spreadsheet, fix unindexed or thin template pages, design hub or pillar pages, decide breadcrumbs or interlinking, brief developers on SEO page builds, or asks how to get pages live and indexed, even if they never say "programmatic SEO".

Complete guide for system automation, DevOps, remote execution, and shell scripting (Bash, Zsh, Fish, Dash, Sh) with Org Babel.

Writes blog posts for your own site, from topic to a review-ready post: picks topics from real search data, writes the brief, checks every fact, drafts in your voice, makes a 1200x630 featured image and puts the post in the site's content folder with frontmatter the build accepts. Reads BLOG.md at the project root first for voice, audience, banned words, categories, cadence and the frontmatter contract. Use whenever the task is to write, draft, research, brief, refresh, update, schedule or publish a blog post or article, plan what to write next, find keywords or topics for a blog, fix an old post that Google shows but nobody clicks, or fill in BLOG.md. Triggers on: write a blog post, blog idea, blog topics, blog brief, keyword research for the blog, refresh a post, schedule a post, publish the post, BLOG.md, posts/ folder. Social posts and newsletters belong to other skills.

Sets up, from nothing, a remote Linux "devbox" where the Claude desktop app runs its Claude Code sessions over SSH, on the user's Claude subscription (no API credits), then mirrors the user's local Claude environment there (global CLAUDE.md, skills, memory, plugin hooks, projects). Includes a shared Chrome logged in to the user's accounts that agents drive through MCP, a virtual screen viewable in a web browser, long tasks in tmux, and automatic continuation of sessions when the user's computer is turned off. Use when the user wants Claude Code sessions to run on a VPS or another machine instead of their laptop (low RAM, many sessions, Chrome tabs, installs), asks which server to rent for it, wants sessions to keep going with the computer off, wants a browser with their accounts on a server, wants to copy their CLAUDE.md, skills or memory to a remote host, or wants to repair, update or verify an existing devbox.

Author Roblox character animations (R15 or R6) headlessly in Blender from a JSON pose spec, verify them by rendering contact sheets and GIFs and match a reference pose. Use this whenever the user wants to make, generate, or automate a Roblox animation — a punch, swing, emote, idle, walk, death, tool action — or wants an animation built to match a reference pose, or asks to animate a .blend rig without opening Blender's UI. Also trigger when a Roblox animation "comes apart", limbs detach, a rig poses correctly in script but renders wrong, limbs pop or flip between frames, or a pose looks nothing like the reference.

Write or refresh Google Ads responsive search ads (RSAs): full sets of truthful headlines and descriptions built from the real offer, buyer and landing page, deliberate pins only for required messages, character-validated fields and a paused Google Ads Editor upload file. Use for Google Ads ad copy, RSA builds, new headlines and descriptions, ad copy refreshes and RSA reviews. Read-only: it never changes the account. Not for keyword structure, negative keywords, sitelinks or other assets, bidding or full account audits.

Produces senior-level commit messages and change descriptions for any Git-managed work, including software development, Infrastructure as Code (IaC), cloud platforms, Kubernetes, GitOps, CI/CD, DevOps, security, observability, data platforms, documentation, architecture, operations, and configuration management. Use whenever creating commits, reviewing changes, preparing pull requests, or summarizing a branch before merge.