Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.
Add full Arabic and right-to-left support to a web app (Next.js App Router and Tailwind first), covering locale routing, dir, logical CSS, mirroring, Arabic typography, numbers, dates, plurals, mixed-direction text, forms, translation and visual QA. Use when the user asks to add Arabic, RTL, or localize an app for Arabic speakers. Don't use for translating a single document.
Map a project's architecture from its code and config and generate one interactive, offline web page that shows each component with its technology icon, animated request flows between them, and possible risks such as a single database or missing rate limiting. Use when the user asks to visualize, map or explain a system's architecture or request flow. Don't use for UI mockups, live production monitoring or code-level call graphs.
Strict low-cost mode for the rest of the session, with tighter reading limits and terse replies. Slash command only. Run it when you want the agent to spend as few tokens as possible.
Apply clear-code rules while writing or changing code, then run the project's own lint, type check and tests and fix what fails. Use when writing or editing code, when asked to clean up or tidy code, and before saying a coding task is done. Do not use for prose, design work or one-line fixes the user dictated exactly.
Write a clear git commit for the current changes, splitting unrelated work into separate commits. Use when asked to commit or to split work into logical commits. Do not use to push, amend or rewrite history unless the user asks.
Plan and write safe database migrations with reversible SQL, backfills and lock-time warnings, checked by a linter. Use when the user asks to change a table, add or drop a column or index, rename something or backfill data in an existing database. Do not use for designing a new schema from scratch or tuning queries.
Speed up a slow SQL query by reading its execution plan, finding missing indexes and N+1 patterns, and proving the change with before and after timings. Use when the user reports a slow query, a slow page caused by the database, or asks to review query performance. Do not use for schema design or migrations.
Design or review a relational database schema from requirements, producing normalized tables, keys, constraints and indexes as a migration plus a short ER summary. Use when the user asks to design tables, model data, add an entity or review a schema. Do not use for query tuning, data migration of existing rows or non-relational stores.
Generate realistic, deterministic seed or fixture data for a database from a short spec, as SQL inserts, JSON or CSV, with a seed so reruns match. Use when the user needs sample data, test fixtures or a populated development database. Do not use to copy or anonymize production data.
Debug a bug with hypotheses instead of guesses. Reproduce it, isolate it, verify the root cause, make the smallest fix and add a regression test. Use when the user reports a bug, an error, a crash or unexpected behavior and the cause is unknown. Do not use for planned features or refactors.
Build or extend a clean, minimal startup or SaaS website (landing page, pricing, about, contact) using this skill's component kit and design tokens. Use when the user asks for a minimal, modern product or startup site, or adds pages to a site already using this kit. Do not use for animation-heavy portfolios, dashboards, docs sites or dark editorial styles.
Build or extend an animation-heavy, dark, editorial portfolio website (kinetic type, smooth scroll, scroll reveals) using this skill's component kit and design tokens. Use when the user asks for a motion-rich portfolio, studio or personal site in this style, or adds pages to a site already using this kit. Do not use for dashboards, docs sites, or minimal/corporate styles.
Tighten and clarify writing without changing its meaning or voice. Use when asked to edit, proofread or shorten an article, email, README or other document. Do not use to write new content or to change what an argument says.
Map an unfamiliar codebase before changing it, reading only what the question needs. Use when asked to understand, survey or find where something lives in a repository, or before changing code you have not seen. Do not use when you already know where the change goes.
Write a short handoff note so a fresh session can resume cheaply, and read it at the start of the next one. Use when stopping work with something unfinished, when a session is getting long, or when asked to hand off. Do not use to record a finished task or to document the code.
Carry out an agreed plan in small steps, checking each one before the next. Use after a plan is approved, or when asked to build or change something with clear requirements. Do not use when the requirements are still unclear; plan or ask first.
Turn a request into an ordered, checkable plan before touching any files. Use when a task spans several files or steps, when requirements are unclear, or when asked to plan first. Do not use for one-step changes or for questions that need an answer rather than a plan.
Review a diff for correctness, clarity and risk, and report findings by severity. Use when asked to review changes, a pull request or a commit. Do not use to write the fix; report first, change code only when asked.
After editing UI code in a project that supports right-to-left languages (it has i18n.config.json with an Arabic locale), check the changed files for RTL problems such as physical Tailwind classes, hardcoded dir="ltr", unflipped directional icons and locale-less number or date formatting. Read-only. Use when you have just written or changed React, Next.js or CSS files in such a project. Don't use in projects without RTL support.
Fix failing tests from trimmed runner output, find the real cause, and never weaken an assertion without asking. Use when tests fail, CI is red, or the user asks to fix failing tests. Do not use to write new tests or to make a red suite green by deleting or skipping tests.
Write tests only for the code that changed, using the project's own test framework and conventions, then run them. Use when the user asks to add tests, or after changing logic that has no tests. Do not use to rewrite a whole test suite or to chase coverage numbers.
Run the project's lint, type check and tests (and the route JavaScript budget when available) and print failures only, before calling work done. Use when finishing a coding task, before a commit or pull request, or when asked to check that everything still passes. Do not use as a substitute for reading the failures; fix what it reports.
Keep a Next.js site fast while building or changing pages, components, images, fonts and animation, and check each route's JavaScript against a budget. Use when adding or editing pages, media or motion in a Next.js App Router project, or when asked to speed a site up. Do not use for backend-only work, non-Next.js stacks or visual design decisions.
Apply web security rules while building or changing a Next.js app (input validation, authorization, secrets, headers, redirects, HTML, rate limits, dependencies) and run secret, header and dependency checks. Use when touching forms, server actions, route handlers, auth, data access, environment variables, next.config or dependencies, or when asked for a security review. Do not use for non-web code or for penetration testing a live site.