Reusable AI skill packs organized by library and topic.
This repository also configures its own agents. See what agent-skills tells them →
Skill Claude CodeCodex
Author and consume next-intl messages with ICU, rich text, typed arguments, validation, and Intl formatting across Next.js components. Use when tasks mention useTranslations, useFormatter, message catalogs, t.rich, t.markup, numbers, dates, lists, relative time, message validation, or extracted messages.
Skill Claude CodeCodex
Set up and maintain next-intl locale routing and navigation in Next.js App Router. Use when tasks mention defineRouting, createMiddleware or proxy.ts, createNavigation, locale prefixes, localized pathnames, domains, locale switchers, alternate links, locale cookies, or CMS-driven localized URLs.
Skill Claude CodeCodex
Configure next-intl request runtime and server-side integration in Next.js App Router. Use when tasks mention i18n/request.ts, getRequestConfig, requestLocale, setRequestLocale, getTranslations, NextIntlClientProvider inheritance, Server Components, Server Actions, metadata, route handlers, sitemap, manifest, OG…
Skill Claude CodeCodex
Set up next-intl tooling and durable localization workflows for Next.js. Use when tasks mention createNextIntlPlugin, TypeScript augmentation, createMessagesDeclaration, useExtracted, getExtracted, precompile, message loading strategies, missing-message handling, Crowdin, or next-intl migration and upgrade decisions.
Skill Claude CodeCodex
Fetch data and control cache behavior in Next.js App Router. Use when tasks mention fetch in Server Components, request memoization, cache components, use cache, unstablecache, revalidatePath, revalidateTag, updateTag, generateStaticParams, route-segment cache behavior, or deciding whether a route should stay static…
Skill Claude CodeCodex
Manage metadata and SEO surfaces in Next.js App Router. Use when tasks mention metadata, generateMetadata, generateViewport, ImageResponse, Open Graph or Twitter images, robots.txt, sitemap.xml, manifest.json, JSON-LD, canonical URLs, or deciding between metadata exports and metadata file conventions.
Skill Claude CodeCodex
Set up Next.js App Router, choose the right starting path, and route work to the correct follow-up Next.js skill. Use when tasks mention create-next-app, App Router installation, project structure, layouts, pages, linking, navigation, Server vs Client Components, path aliases, src/app, App Router vs Pages Router…
Skill Claude CodeCodex
Use when a Next.js App Router route is slow because of render cost, client bundle weight, hydration work, or weak server-client boundaries. Covers server-first rendering, client boundary minimization, route cost diagnosis, and avoiding fashionable but expensive patterns.
Skill Claude CodeCodex
Choose rendering mode and request runtime boundaries in Next.js App Router. Use when tasks mention Server vs Client Components, static vs dynamic rendering, streaming, route segment config, runtime = 'edge', Node.js vs Edge decisions, or legacy middleware questions that now map to proxy.ts.
Skill Claude CodeCodex
Build safe mutation and API boundaries in Next.js App Router. Use when tasks mention use server, Server Functions, Server Actions, Route Handlers, route.ts, form action, formAction, cookies, redirects after mutations, request and response handling, or deciding between UI-driven mutation flow and explicit HTTP…
Skill Claude CodeCodex
Use when a Node.js API is slow because of per-request work, bad concurrency shape, I/O amplification, or weak throughput defaults. Covers p95-first analysis, request path simplification, batching, cancellation, and avoiding fake optimizations that hurt operability.
Skill Claude CodeCodex
Use when building or debugging Node.js applications and libraries. Covers modules, async flows, streams, testing, shutdown, environment handling, logging, performance, and modern TypeScript execution choices in Node.js.
Skill Claude CodeCodex
Use when someone claims a performance improvement or worries about a regression and needs proof. Covers before-versus-after comparison, stable verification scenarios, user-visible metrics, and guarding against “faster on my machine” conclusions.
Skill Claude CodeCodex
Use when something is slow but the owning bottleneck is still unclear. Covers choosing the right performance symptom, assigning likely layer ownership, and routing optimization work before the model starts changing random code.
Skill Claude CodeCodex
Use when someone wants to optimize code but the team has not profiled it yet. Covers baseline capture, profiling representative scenarios, finding dominant hotspots, and refusing performance folklore as a substitute for measurement.
Skill Claude CodeCodex
Use when reusing authenticated state or modeling multiple user roles in Playwright without leaking state between tests. Covers storage state, setup projects, browser contexts, role-based projects, and the rule that isolation should stay intact even when login is optimized away.
Skill Claude CodeCodex
Use when a Playwright test is flaky, failing only in CI, or needs investigation rather than another retry. Covers traces, inspector, UI mode, codegen cleanup, and turning intermittent failures into concrete synchronization or contract fixes.
Skill Claude CodeCodex
Choose a stable Playwright test architecture before writing more tests. Use when tasks mention Playwright adoption, test folder layout, fixture boundaries, playwright.config.ts, projects, retries, parallelism, CI posture, browser coverage, or deciding which Playwright follow-up skill should own the work.
Skill Claude CodeCodex
Use when writing or fixing Playwright tests that should stay stable against UI refactors. Covers locators, actionability, web-first assertions, accessible roles, text and label contracts, and the rule that tests should bind to user-visible semantics before DOM structure.
Skill Claude CodeCodex
Use when Playwright tests need stable synchronization with backend behavior or deliberate network control. Covers request waiting, route interception, mocking boundaries, API request contexts, and the rule that tests should not over-mock what they actually need to verify.
Skill Claude CodeCodex
Use when React performance problems come from rerender scope, bad state ownership, or cargo-cult memoization. Covers state placement, context blast radius, render cost boundaries, and choosing fixes that reduce work instead of hiding it.
Skill Claude CodeCodex
Use when the real security question is whether the current actor is allowed to do the thing, not whether they are merely authenticated. Covers server-side authorization, ownership checks, tenant boundaries, role use, and avoiding trust in user-controlled identifiers.
Skill Claude CodeCodex
Use when a web app risks exposing secrets or privileged behavior across the client-server boundary. Covers server-only logic, public env variables, scoped tokens, browser trust assumptions, and moving sensitive work out of the client.
Skill Claude CodeCodex
Use when reviewing a web app for practical security risk before coding or shipping. Covers trust boundaries, attacker-controlled input, privileged actions, secret exposure, and routing the problem to narrower security skills without pretending to do a full security audit.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: