Writing & Editing

Make any prose an AI writes read like a careful human wrote it - clear, direct, specific, and persuasive. Use when drafting, editing, or reviewing emails, memos, executive summaries, reports, proposals, documentation, posts, slide text, or any writing meant for people, and whenever text sounds generic, padded, promotional, or "AI-like." Covers audience and purpose, top-down structure, plain style, and a full checklist of AI-writing patterns to remove.

Detect and avoid the telltale patterns of AI-generated (LLM, chatbot, ChatGPT-style) writing, based on Wikipedia's "Signs of AI writing" field guide by WikiProject AI Cleanup. Use this skill whenever the user asks whether text was written by AI, wants a draft checked for "AI tells", "slop", puffery or generic chatbot prose, is reviewing Wikipedia edits, articles, essays, submissions or comments for undisclosed AI use, or wants writing revised so it reads as specific and human. Also consult it before writing or editing encyclopedic, journalistic, academic or other formal prose, where inflated significance, vague attributions, invented sources or chatbot formatting would hurt the result.

Write or rewrite a README that makes a stranger star the repo within 15 seconds. Use when creating a README, when the user says the README is weak, before any launch, or when a repo gets traffic but few stars.

Write or rewrite selling copy, then humanize it so it reads like a person wrote it. Use when someone asks to make copy better, write a hook, headline or subject line, or write or rewrite a cold email, DM, follow-up, sales page, landing page, social post, video script, flyer, ad, sales script or offer. Uses notes from Cialdini, Bly, Hormozi, Harry Dry, Tyson 4D and Stefan Georgi, then runs a humanizer pass last.

Apply eight storytelling principles to business communication: predictive processing, establishing shots, word pictures, story structure, Chekhov's gun, a specific villain, sugar coating, and bookending. Use for writing or improving emails, posts, landing pages, scripts, sales narratives, and content strategy; also for discovering what existing content to edit, remove, or develop. Adapt the method to the business, audience, and request; answer direct factual questions plainly.

Write or rewrite a short, plain-English pull request description from the actual diff and current PR state. Use when a PR body needs to explain the problem, why the change is needed, and what is true after merge.

When the user wants to edit, review, or improve existing marketing copy, or refresh outdated content. Also use when the user mentions 'edit this copy,' 'review my copy,' 'copy feedback,' 'proofread,' 'polish this,' 'make this better,' 'copy sweep,' 'tighten this up,' 'this reads awkwardly,' 'clean up this text,' 'too wordy,' 'sharpen the messaging,' 'refresh this content,' 'update this page,' 'this content is outdated,' or 'content audit,' 'this sounds like AI,' 'AI slop,' 'de-slop this,' or 'make it sound human.' Use this when the user already has copy and wants it improved or refreshed rather than rewritten from scratch. Every edit removes AI tells such as 'it's not X, it's Y' reveals, 'no X, no Y, no Z' lists, and sentences that trail into extra comma clauses. A request to fix one line gets a point edit that changes only that line. For writing new copy, see copywriting.

Use when the user wants to coordinate AI employees across ChatGPT or Codex projects, define controller and executor roles, write bounded task cards, or review execution receipts and evidence before acceptance.

Decompose a problem, feature, or goal into candidate ADRs and plan items for a documentation-led repo — one decision per ADR, dependency edges, suggested ordering. Proposes drafts for review and writes nothing until approved. Use when the user says "brainstorm ADRs", "break this down into decisions", "what ADRs do we need for X", "plan out the work", or invokes /brainstorm.

Scaffold or retrofit documentation-led conventions (AGENTS.md, CLAUDE.md, CONVENTIONS.md, ADR catalogue, plan/ queue, _agent/ coordination) into a repo. Use when the user asks to "set up conventions", "bootstrap ADRs", "scaffold the documentation-led layout", or "add AGENTS.md and a plan queue". Works on fresh repos and existing ones — preserves existing content and merges rather than overwrites. Opens with an express / guided / full depth choice, so a quick conservative setup needs almost no questions.

Write or extend a conformance suite from a CONTRACT*.md and published sources only, by an author who never sees an implementation, kept read-only while implementations are built against it. Use when a contract has or will have several implementations, or a wrong expected value would be invisible to the implementer, as in cryptography, codecs, and wire protocols.

Fill-in Markdown templates for integration project deliverables (DAT architecture document, DEX operations manual, production release sheet, interface contract, flow matrix, RACI, RAID log, UAT sign-off) and the writing rules that apply to all of them. Load when an agent must produce, update or review a project deliverable.

Write an Architecture Decision Record (ADR) or a target-architecture note from a question or context - options, knock-out and weighted criteria, trade-offs, decision, consequences, reversibility and compliance impacts (security/accreditation, GDPR, DORA, sovereignty). Use when a structural technical choice must be made, documented or reviewed, or when a perimeter needs a documented target state.

Open a live office view of every Claude Code, Codex and Gemini CLI agent on this machine — who is thinking, running a tool, writing a reply, waiting for the user, or done. Use when the user types /myoffice, asks what their agents are doing, wants to see all running sessions, or which agent is stuck or finished ("我的 agent 在幹嘛", "show my agents").

Create or maintain project documentation such as README.md, AGENTS.md, status.md, and llms.txt, or audit documentation for drift after code changes. Use for documentation updates, project status snapshots, and explicit docs-sync checks; do not use for Git commits, GitHub repository governance, or unrelated long-form writing. Also use when a work phase completes, pauses, blocks, or is handed off, even if the session is mainly code work.

Write, revise, or audit OxideTerm tests for observable behavior, credible regressions, and independent contracts. Use when diagnosing a failed test, reviewing test quality, or consolidating redundant coverage. Ordinary test changes use a focused check, not a repository-wide audit.

Document a research method as a METHOD.md with provenance and validation sections, in the format used by the research-methods repository. Use when asked to write up, record, or formalize a method, procedure, algorithm, or analysis recipe for reuse.

Interview the user in depth about a feature, then write a self-contained SPEC.md to implement from in a fresh session. Use before building anything large.

Draft, revise, and finalize Chinese natural science fund proposals in the user's established academic style. Use when Codex must edit a grant DOCX, learn style from a reference fund PDF/DOCX, restructure research rationale, literature review, scientific questions, three research-content blocks, methods, limitations, feasibility, annual plans, expected outcomes, achievements, sensors, citations, or references; synchronize正文 with figures; verify publication and bibliography facts; or perform final logic, formatting, and submission-readiness audits.

英語で書かれた学術論文・原稿のPDFを、プロの英文校正者として校正するスキル。学術出版ガイド（Olson, Guide to Academic and Scientific Publication）に基づくチェックリストで、英文法・語法、US/UK綴り、略語、句読点、時制、代名詞、用語の一貫性、図表、引用と文献リスト、タイトル・要旨・キーワード、投稿規定への適合を確認し、修正案つきのレポートと赤入れ（注釈付き）PDFを作る。モードは2つ：「1回校正」（指定したPDFを1回校正）と「loop」（校正→原稿に反映してPDFを再生成→再校正を、指摘が無くなるまで繰り返す）。使い方は `/academic-writing-loop 1回校正 paper.pdf` または `/academic-writing-loop loop paper.pdf`。英語論文のPDFについて「英文校正して」「英語をチェック／添削して」「ネイティブチェック」「文法を直して」「proofreadして」「投稿前に英語を磨きたい」などと頼まれたら、スキル名が出ていなくても使うこと（内容の査読は paper-review-loop、技術的な誤りの検証は deep-review-modoki の担当）。Use for English proofreading/copyediting of academic paper PDFs, single pass or iterative proofread→revise→re-proofread loops.

Create or update multiple CMS items in a Webflow collection with validation and diff preview. Use when adding multiple blog posts, products, or updating fields across many items.

Use this skill when committing code, writing commit messages, or reviewing git history in this Next.js project. Triggers on: "commit this", "write a commit message", "what should my commit say", "git commit", or any request to stage and commit changes. Provides project-specific conventional commit format, Thai-language comment conventions, and scope rules matched to this codebase's structure.

Write and organize unit tests for functions, methods, and classes using `package:test`. Use when creating new logic or fixing bugs to ensure code remains correct and regression-free.

Import an external astronomical catalog into HATS, verify it against the source, document it, host it and register it on data.lsdb.io, following the LSDB team's conventions. Use when the user says "create a HATS collection from X", "import X into HATS", "re-import X", "put X on data.lsdb.io", "make X available like <an existing catalog>", or "verify the HATS copy of X against the source". NOT for Rubin Butler products (they have their own pipeline) and NOT for catalogs already in HATS that only need a page or a mirror (use reference/publishing.md directly).