Claude Code skills
183 of them run by themselves.
Loops that turn ideas into specs, specs into tickets and tickets into reviewed pull requests.
Set up Habit: follow the instructions in https://gethabit.dev/LOOP.md
- 1Copy
- 2Paste into Claude Code, Codex or Cursor
- 3Your agent finds the right skill
Improve Codebase Architectureby Matt Pocock · 1.2M installsFinds messy parts of your code and helps you make them simpler.↻Code that gets easier to change each week. Less time lost on hard-to-test spots.
agent-browserby vercel-labs · 1.2M installsLets your agent use a web browser to click, type and look at pages.↻Your web app tested every night. Bugs found early, so fewer reach your users.TDDby Matt Pocock · 1.1M installsTeaches your agent to write a test first, then the code.↻Features built test-first, one small step at a time. Fewer bugs and safer changes.
Google Agents CLI Scaffoldby google · 506K installsThe google agents cli scaffold skill by Google creates, enhances and upgrades agent projects with agents-cli scaffold: templates, deployment targets, CI/CD runners, session storage and the prototype-first pattern.↻Agent projects stay on current templates.
Caveman Commitby Julius Brussee · 411K installsThe caveman commit skill writes short Conventional Commits messages that state intent only: imperative subject, allowed type and optional scope, a body only when needed, no filler.↻Every commit message short and exact.
Prisma CLIby prisma · 381K installsThe prisma cli skill is a reference for the Prisma ORM 7 command line: init, generate, migrate, db, dev, studio and mcp, with a consent checkpoint before destructive commands.↻Schema valid, migrations in sync.Prisma Driver Adapter Implementationby prisma · 369K installsThe prisma driver adapter implementation skill is a reference for building Prisma ORM 7 SQL driver adapters: transaction lifecycle, savepoints, result mapping, error preservation and a verification checklist.↻Driver adapter changes pass the checklist.Cavecrewby Julius Brussee · 355K installsCavecrew is a skill that tells an agent when to hand work to three compressed subagents: an investigator to find code, a builder for one or two file edits and a reviewer for diffs.↻Small fixes with lean context.
Supabaseby supabase · 338K installsGuides your agent through Supabase work like data, login and safety rules.↻Safer database changes every time. Fewer security holes and fewer broken updates.
Systematic Debuggingby Jesse Vincent · 304K installsMakes your agent find why a bug happens before it fixes it.↻Bugs fixed at the source, not patched over. Fewer bugs come back.Requesting Code Reviewby Jesse Vincent · 260K installsThe requesting code review skill has your agent send its diff to a fresh reviewer subagent with clean context, then fix what it finds by severity.↻Every task reviewed before merge.Test-Driven Developmentby Jesse Vincent · 257K installsMakes your agent write a failing test first, then the least code to pass.↻Code you can trust. Fewer bugs, since every change starts with a test.
Fixby react · 243K installsThe fix skill has your agent run yarn prettier and yarn linc before a commit and report lint errors that need manual work.↻Commits that pass lint checks.
Firebase Security Rules Auditorby firebase · 137K installsFirebase security rules auditor is a skill that red-teams Firestore and Cloud Storage rules for bypasses, privilege escalation, missing limits and type checks, and returns a 1 to 5 score in JSON.↻No rules change ships unaudited.
Durable Objectsby cloudflare · 107K installsThe durable objects skill helps agents build, debug and review Cloudflare Durable Objects code for persistent state and coordination, with current docs, wrangler config, storage, alarms and Vitest testing.↻Durable Object code stays safe.
Convexby convex · 107K installsHelps you build and check your Convex backend.↻Every backend change is checked for bugs and leaks. Fewer bugs reach your users.Agents SDKby cloudflare · 103K installsHelps your agent build and check Cloudflare agent apps.↻Setup mistakes are caught on every change. Fewer bugs reach your live app.
Playwright Stealth Verifyby liarjsdev · 98K installsThe playwright stealth verify skill checks whether a Playwright, Puppeteer, Selenium or CDP browser shows a coherent fingerprint, using liarjs checkPage on a Page you already have.↻Automation harness stays coherent per PR.
Solana Devby solana-foundation · 95K installsThe solana dev skill is an end-to-end playbook for Solana: wallet connection, Anchor and Pinocchio programs, Kit clients, Surfpool and LiteSVM testing, security review and v1 transactions.↻Solana changes built, tested, risk-noted.
Nx Workspaceby nrwl · 73K installsThe nx workspace skill helps your agent explore an Nx monorepo read-only: list and filter projects, read full resolved targets, check nx.json and query the project graph, for teams who work in Nx workspaces.↻A current map of your monorepo.
Code Review and Qualityby Addy Osmani · 70K installsThe code review and quality skill gives your agent a five-axis review for correctness, readability, architecture, security and performance, with severity labels and test checks, for teams who review code from people or agents.↻Every PR reviewed before merge.
GSAP Performanceby greensock · 67K installsThe gsap performance skill is the official GreenSock guide for smooth GSAP animations: transforms over layout properties, careful will-change, batched reads and writes, stagger, quickTo and cleanup.↻Keep GSAP animations smooth.GSAP Reactby greensock · 64K installsThe gsap react skill is the official GreenSock guide to GSAP in React and Next.js: the useGSAP hook, refs and scope, gsap.context() with cleanup, contextSafe callbacks and server rendering.↻No leaking GSAP animations in React.
InsForgeby InsForge · 59K installsThe insforge skill guides your agent through app code with @insforge/sdk: database CRUD, auth, storage, functions, AI through OpenRouter, realtime, email and Stripe or Razorpay payments.↻Catch InsForge SDK mistakes early.
Ponytail Reviewby DietrichGebert · 54K installsThe ponytail review skill makes your agent review a diff like the developer on call: bugs, security, real load, missing tests, speed and code to delete, each with a concrete case and the smallest fix.↻Every change reviewed before merge.
shadcn/uiby google-labs-code · 53K installsHelps you add and tune ready-made parts for your website screens.↻New screen parts get checked every time. Fewer bugs and a cleaner site.
Golang Code Styleby samber · 44K installsThe golang code style skill gives Go developers clear style rules that linters cannot judge: line breaking, variable declarations, control flow, function design, comments and file layout.↻Go code stays clear, not clever.Golang Testingby samber · 43K installsThe golang testing skill helps Go developers write and review production-ready tests: table-driven tests, testify, parallel tests, fuzzing, goleak leak checks, synctest, coverage and integration tests with build tags.↻Go tests stay fast, isolated, leak-free.Golang Error Handlingby samber · 43K installsThe golang error handling skill teaches Go developers idiomatic error handling: wrapping with %w, errors.Is and errors.As, errors.Join, sentinel and custom errors, panic and recover, and structured logging with slog.↻No swallowed or double-logged errors.Golang Design Patternsby samber · 42K installsThe golang design patterns skill guides agents to idiomatic Go: functional options, no init() or global state, safe enums, timeouts, retries, streaming and a fitting architecture.↻Go code stays idiomatic and bounded.Ponytail Debtby DietrichGebert · 40K installsPonytail debt collects every shortcut: comment in a repo into one debt ledger with its limit and upgrade trigger, and flags markers with no trigger.↻No shortcut quietly becomes permanent.
Limrun iOS Simulatorby limrun-inc · 39K installsThe limrun ios simulator skill drives an app on a Limrun cloud iOS simulator: tap, type, element tree, logs, screenshots, video, tunnels, proxy, camera and Face ID.↻Each build gets a smoke test.
Flutter Add Widget Testby flutter · 39K installsThe flutter add widget test skill writes component tests with WidgetTester, finders and matchers to check rendering, taps, scrolling and text input.↻Every new widget gets a test.
SwiftUI Expert Skillby AvdLee · 35K installsHelps write and check SwiftUI apps for Apple devices.↻Every code change is checked for old and wrong code. Fewer bugs.
Sanity Best Practicesby sanity-io · 31K installsThe sanity best practices skill gives Sanity developers topic guides for schemas, GROQ, TypeGen, Visual Editing, Portable Text, localization, migrations, Functions and framework integrations.↻Sanity changes follow house rules.
Migrate Radix to Baseby shadcn-ui · 29K installsThe migrate radix to base skill moves shadcn wrappers and React code from Radix UI to Base UI step by step, with golden pairs, typechecks and a report per component.↻One component migrated per run.
Review PRby warpdotdev · 29K installsReviews a code change and writes notes on the exact lines.↻Every code change gets a checked review. Fewer bugs reach your users.Fix Errorsby warpdotdev · 28K installsThe fix errors skill helps developers fix compile errors, clippy and fmt failures, WASM errors and test failures in the warp Rust codebase with the narrowest useful command.↻Red builds turned green, narrowly.
Extension Email Marketingby caffeinelabs · 28K installsThe extension email marketing skill helps Caffeine AI app builders add marketing email to subscribers, with topics, email verification and a working unsubscribe page.↻Every marketing mail can unsubscribe.Extension Object Storageby caffeinelabs · 27K installsThe extension object storage skill helps Caffeine AI app builders store images, videos and files off-chain with on-chain references, using ExternalBlob on backend and frontend.↻File uploads keep working.
Vue Debug Guidesby vuejs-ai · 26K installsThe vue debug guides skill gives Vue 3 developers an index of symptoms with matching fixes for runtime errors, warnings, async failures and SSR hydration bugs.↻Vue warnings get fixed, not ignored.
Argent React Native Optimizationby software-mansion · 26K installsThe argent react native optimization skill helps React Native developers speed up apps by measuring first, then running lint, semantic and profiling sweeps with one fix per cycle.↻The app gets measurably faster.Argent React Native App Workflowby software-mansion · 26K installsThe argent react native app workflow skill guides agents to start, debug, rebuild and test React Native apps on iOS simulators and Android emulators with argent tools.↻The app builds and runs clean.Argent React Native Profilerby software-mansion · 26K installsThe argent react native profiler skill helps React Native developers measure re-renders and CPU hotspots on Hermes apps, produce ranked reports and compare before and after a fix.↻Render regressions show up early.
Gemini API Devby google-gemini · 25K installsThe gemini api dev skill helps developers write Python and TypeScript code for the Gemini API, with current models, the interactions API, streaming, agents and migration notes.↻No deprecated Gemini models ship.Sanity Migrationby sanity-io · 22K installsThe sanity migration skill plans, builds and reviews moves into Sanity from AEM, Contentful, Strapi, Webflow, WordPress, Payload, Drupal or Markdown, with Portable Text conversion, asset migration, redirects and validation.↻Migration reruns converge, checks pass.
Next Cache Components Adoptionby vercel · 20K installsThe next cache components adoption skill turns on Cache Components in a Next.js App Router app and works through blocking routes feature by feature until next build passes.↻One feature adopted per run.
Golang Proby Jeffallan · 19K installsThe golang pro skill helps agents write idiomatic Go: concurrency with goroutines and channels, gRPC or REST microservices, generics, error handling, pprof profiling and table-driven tests.↻Go changes that pass every gate.
agent-deviceby callstack · 18K installsTaps through your app on a phone or TV like a real user.↻Every new build gets a real test. Broken screens are found the day they appear.Next Cache Components Optimizerby vercel · 18K installsThe next cache components optimizer skill helps Next.js teams make routes paint instantly, using a failing instant() test that turns green and then guards against regressions.↻One more route paints instantly.
OpenCLI AutoFixby jackwener · 18K installsThe opencli autofix skill repairs broken OpenCLI adapters when a command fails: it collects a trace, patches the adapter, retries, and files an upstream issue after a verified fix.↻Broken adapters fixed and retried.
LangChain Fundamentalsby langchain-ai · 17K installsThe langchain fundamentals skill shows developers how to build LangChain agents with create_agent, define tools with @tool or tool(), add middleware for human approval and error handling, and get structured output.↻LangChain agents built the current way.LangGraph Human in the Loopby langchain-ai · 17K installsThe langgraph human in the loop skill shows developers how to pause a LangGraph graph with interrupt(), resume it with Command(resume=...), build approval and validation flows, and keep side effects safe on resume.↻Paused graphs resume without duplicates.Swift Concurrencyby AvdLee · 17K installsFixes timing errors in your Swift app code.↻Your move to Swift 6 gets done in small safe steps. Fewer bugs, no big risky change.Electronby vercel-labs · 17K installsLets your agent use a desktop app built with Electron.↻Every new build gets a quick test. You find broken screens before users do.OpenCLI Adapter Authorby jackwener · 16K installsThe opencli adapter author skill helps your agent build or improve OpenCLI JS adapters from observed site behavior, existing adapters and browser network evidence, then check and test them.↻Broken site adapters get fixed fast.
Flutter Animationsby madteacher · 15K installsThe flutter animations skill helps Flutter developers add, fix, refactor and debug motion: implicit and explicit animations, Hero transitions, staggered timing, physics, curves and reduced-motion support.↻Clean Flutter motion, analyzer approved.
Fix Security Vulnerabilities with Strixby usestrix · 15K installsThe fix security vulnerabilities with strix skill helps your agent patch findings from a Strix pentest at the root cause and re-run Strix to prove each fix works.↻Findings fixed and proven closed.
Huntby tw93 · 15K installsThe hunt skill finds the root cause of an error, crash or regression before any fix, then proves the fix with a repro and a regression test.↻Find the cause before fixing.
Install Anti-Slopby dmmulroy · 13K installsThe install anti-slop skill installs, updates or migrates the vendored anti-slop Oxlint plugins in a repository, enables their rules and records provenance, for teams who want stricter lint rules against sloppy code.↻Lint rules stay current, customizations kept.
Email best practicesby resend · 13K installsThe email best practices skill by Resend guides developers through deliverable, compliant and accessible email: SPF, DKIM and DMARC, transactional versus marketing mail, consent, retries, webhooks, bounce handling and screen reader friendly templates.↻Emails stay deliverable and accessible.
Svelte code writerby sveltejs · 13K installsThe svelte code writer skill gives your agent the @sveltejs/mcp CLI to look up Svelte 5 and SvelteKit docs and run the svelte-autofixer on components, for developers who write or review Svelte code.↻Ship Svelte components the autofixer accepts.React Emailby resend · 12K installsThe react email skill builds HTML email templates with React components, renders them to HTML and plain text, embeds the visual email editor and sends with Resend.↻Email templates render in every inbox.
Genshijin Commitby interfacex-co-jp · 12K installsThe genshijin commit skill writes very short Conventional Commits messages in Japanese or English, with the reason first and a body only when it adds context.↻Short commit messages that explain why.Genshijin Reviewby interfacex-co-jp · 12K installsgenshijin review is a skill that writes very short pull request review comments in Japanese: one line per finding with line, problem and fix, for developers who want review signal without filler.↻Short review comments on every PR.
HeroUI Nativeby heroui-inc · 11K installsThe heroui native skill helps React Native developers build mobile UIs with HeroUI Native components, Uniwind styling and HSL themes, and fetches current component docs before any code is written.↻HeroUI Native code stays native.
CodeRabbit Autofixby coderabbitai · 10K installsThe coderabbit autofix skill helps developers work through CodeRabbit review threads on their current PR: it checks each issue against the code, asks before every fix, and makes one commit and one summary comment.↻Review threads fixed with approval.
Fallowby fallow-rs · 10K installsThe fallow skill gives an agent codebase intelligence for TypeScript and JavaScript: changed-code risk, dead code, duplication, circular dependencies, complexity hotspots and architecture boundaries, all as machine-readable JSON.↻Every PR gets a risk verdict.Supabase Serverby supabase · 10K installsThe Supabase Server skill (supabase-server) guides an agent writing server-side code with @supabase/server: Edge Functions and backend handlers that pick an auth mode, create Supabase clients and verify inbound auth, and migrate legacy patterns.↻Edge Function auth stays correct.
SwiftData Proby twostraws · 9K installsThe swiftdata pro skill writes and reviews SwiftData code with modern APIs: relationships, delete rules, safe predicates, CloudKit limits and indexing.↻SwiftData bugs caught before merge.
Claude Automation Recommenderby Anthropic · 9K installsThe claude automation recommender skill scans a codebase and suggests the most useful Claude Code hooks, subagents, skills, plugins and MCP servers for it, for developers setting up or tuning their agent workflow.↻Fresh automation ideas every month.
MySQLby planetscale · 9K installsHelps plan safe changes to a MySQL database.↻Every database change is checked with proof. Fewer slow or risky changes.
CrewAI design agentby crewaiinc · 8K installsThe crewai design agent skill helps developers design CrewAI agents: role, goal and backstory, LLM choice, tools, limits, planning, delegation, knowledge sources and guardrails.↻Every agent passes the checklist.
Tauri v2by nodnarbnitram · 8K installsTauri v2 helps you build cross-platform desktop and mobile apps with a Rust backend: commands, IPC with invoke, emit and channels, capabilities and permissions, config and troubleshooting.↻Tauri commands and permissions stay wired.
CI/CD securityby superagent-ai · 8K installsThe ci/cd security skill reviews GitHub Actions workflow YAML for risky triggers, permissions, action pinning and injection patterns, with no tools to install, for maintainers hardening their pipelines.↻Catch risky CI workflow changes.Skill securityby superagent-ai · 8K installsThe skill security skill audits an AI agent skill before you install it, pairing a deterministic offline scanner with a reasoned contract check, for developers who want to know if a skill is safe.↻Vet every new skill first.
Find Bugsby getsentry · 7K installsThe find bugs skill from Sentry reviews local branch changes for bugs, security holes and quality issues with a full diff read and a security checklist, then reports verified findings.↻Find bugs before branches merge.
Playwright Skillby lackeyjb · 7K installsOpens your site in a browser and tests it for you.↻Broken pages are found every night, before your visitors see them. Fewer bugs.
shadcn-svelteby huntabyte · 7K installsThe shadcn-svelte skill helps agents add, update, fix and compose shadcn-svelte components with the CLI, following rules for forms, styling, icons and composition.↻Keep Svelte UI components current.
Printing Press Reprintby Matt Van Horn · 6K installsThe printing press reprint skill regenerates an existing printed CLI under the current Printing Press, carrying prior research, novel features and patches into the new run.↻Regenerate CLIs after machine upgrades.
Rust Skillsby leonardomso · 6K installsThe rust skills skill gives 265 Rust coding rules in 26 categories, from ownership and error handling to async, unsafe, API design and performance, ranked by impact.↻Keep Rust code idiomatic and safe.
Ant Design (antd)by ant-design · 6K installsThe Ant Design (antd) skill teaches agents the Ant Design CLI: query props, demos, tokens and docs offline, lint and analyze project usage, and follow migration guides between antd versions.↻Keep antd usage current and clean.
ccc (CocoIndex Code)by cocoindex-io · 6K installsThe ccc (CocoIndex Code) skill lets an agent run semantic code search over the current codebase with the CocoIndex Code CLI, and keep the index fresh with init and index commands, with language and path filters.↻Keep semantic code search fresh.
marimo Notebookby marimo-team · 6K installsThe marimo notebook skill shows how to write marimo notebooks as Python files: cell structure, script mode, reactivity rules, output rendering, PEP 723 dependencies, marimo check and pytest.↻Keep marimo notebooks running clean.
Antigravity Delegate (agy)by amelnagdy · 6K installsAntigravity Delegate (agy) lets an orchestrator hand a bounded coding task to the Google Antigravity CLI, then review the diff, rerun the gates and commit it itself, with a relay that records result.json.↻Delegate to agy, review, then land.
Building Pydantic AI Agentsby pydantic · 6K installsHelps you build and test AI agents in Python.↻Your agents stay tested and easy to trace. Fewer bugs when you change code.
.NET Run Testsby dotnet · 6K installs.NET Run Tests picks and runs the right dotnet test command for classic, VSTest, bridged MTP or native MTP projects, with correct filters for MSTest, xUnit, NUnit and TUnit plus TRX, dumps and coverage.↻Run the right dotnet test command.
no-mistakesby kunchenguid · 6K installsChecks your work step by step before you ship it.↻Fewer bugs reach your users. Every change passes the same checks.Optimizing EF Core Queriesby dotnet · 5K installsThe optimizing ef core queries skill reviews and fixes Entity Framework Core data access: N+1 queries, premature materialization, index-hostile predicates, tracking, Includes, paging, bulk writes and DbContext overlap.↻Catch slow EF Core queries early.React Native Testingby callstack · 5K installsThe react native testing skill writes and fixes component tests with React Native Testing Library v13 and v14: render and screen, query priority, Jest matchers, userEvent, fireEvent, waitFor and async patterns.↻Keep React Native components covered.
Indexion SDDby trkbt10 · 5K installsIndexion sdd drafts spec-driven development requirements from RFCs and standards, then measures how well the implementation conforms with spec align and spec verify, including CI drift gates and a codex or claude validation loop.↻Implementation stays aligned with the spec.
Next.js Cache Architectureby mohamed-hossam1 · 5K installsNext.js Cache Architecture helps agents design caching in Next.js 16+ App Router: use cache, a tag registry, revalidation utilities, Suspense boundaries, personalized content and migration from unstable_cache.↻Cache tags and invalidation stay correct.Indexion Agent Orientby trkbt10 · 5K installsIndexion agent orient builds a pre-edit structure brief so coding agents learn likely code owners, consumer surfaces and unsafe edit locations before implementing.↻Agents know code owners before editing.Indexion Identityby trkbt10 · 5K installsIndexion identity detects name and content drift in functions, files and folders with indexion identity audit, then plans verified renames, moves or splits.↻Names keep matching what code does.Indexion KGFby trkbt10 · 5K installsIndexion kgf inspects and debugs KGF language specs by showing tokens, parse events and extracted edges for a source file, with known PEG and token pitfalls.↻KGF spec edits keep docs extracted.GHA Security Reviewby getsentry · 5K installsGha security review audits GitHub Actions workflows for exploitable bugs like pwn requests, expression injection, credential theft and supply chain attacks, reporting only traced findings with concrete exploit scenarios.↻Exploitable workflow bugs caught before merge.AXIby kunchenguid · 5K installsAxi defines Agent eXperience Interface standards for CLI tools agents use through the shell: TOON output, minimal schemas, truncation, structured errors, exit codes, session hooks and fast version checks.↻Agent-facing CLIs stay AXI-compliant.
React Code Reviewby giuseppe-trisciuoglio · 4K installsReact code review checks React components, hooks, React 19 patterns, state management, performance, accessibility and TypeScript props, producing severity-ranked findings with code examples.↻React changes reviewed before merge.
Developby jsmastery-pro · 4K installsDevelop builds a feature, UI or backend, from an approved spec: it gates on undecided choices and routes them to architect, then reads the spec and AGENTS.md, builds, and advances the scope.↻Features built only from settled specs.
Database Schema Designerby softaworks · 4K installsDatabase schema designer designs SQL and NoSQL schemas with normalization, indexing, constraints, relationship patterns and reversible migrations built in.↻Schema changes pass the design checklist.Test Anti-Patternsby dotnet · 4K installsTest anti-patterns audits a test file or suite in any language and produces a severity-ranked report of tests that verify nothing, swallowed exceptions, flaky or order-dependent tests, duplication and magic values.↻Tests that prove nothing get caught.Ant Designby ant-design · 4K installsHelps use Ant Design parts the right way.↻Every change is checked against real parts. Fewer bugs from wrong code.ASP.NET Core Web APIby dotnet · 4K installsDotnet webapi guides creating and changing ASP.NET Core Web API endpoints with correct HTTP semantics, sealed record DTOs, OpenAPI metadata, RFC 7807 error handling and .http test files.↻New API endpoints follow HTTP conventions.
Spec Implby klerith · 4K installsSpec impl implements an approved spec: it validates the Approved state in any language, creates a branch named after the spec, and works the plan step by step with pauses to review each diff.↻Approved specs implemented on their branch.
Convex Best Practicesby waynesutton · 4K installsConvex best practices covers the production patterns the convex eslint plugin enforces: validators, index reads, idempotent mutations, avoiding OCC write conflicts, pagination, thin wrappers and error handling.↻Convex changes checked against the rules.Run Nx Generatorby nrwl · 4K installsRun Nx Generator (the nx-generate skill) scaffolds libraries, apps and features in an Nx monorepo: it finds the right generator, reads its source, dry-runs, generates and verifies with build, lint, test and typecheck.↻Scaffolded code that passes checks.
GraphQL Operations Guideby apollographql · 4K installsGraphQL Operations Guide (the graphql-operations skill) guides writing queries, mutations and fragments against the project schema: named operations, variables, lean field selection, fragments and type generation.↻Operations match the schema and lint.
PHP Best Practicesby asyrafhussin · 4K installsPHP Best Practices reviews PHP 8.x code against type safety, modern features, PSR standards, SOLID principles, error handling, performance and security rules, matched to the project's version.↻PHP code checked against modern rules.Xcode Build Fixerby AvdLee · 4K installsXcode Build Fixer applies approved Xcode build optimizations, build settings, script phase guards, source fixes and SPM restructuring, then re-benchmarks to prove the gain.↻Approved build fixes proven by benchmark.Xcode Build Orchestratorby AvdLee · 4K installsXcode Build Orchestrator runs an end-to-end Xcode build speed-up: benchmark, specialist analysis, ranked recommendations, an approval gate and re-benchmarking, delegating fixes to xcode-build-fixer.↻A ranked, approved build speed-up plan.
Alliumby juxt · 4K installsThe allium skill is a formal language and loop for specifying software behaviour: it captures intent in .allium files, generates integration and end-to-end tests, and drives spec, tests and code to agreement.↻Spec, tests and code all agree.
Open Code Review Delegateby alibaba · 4K installsThe open code review delegate skill uses the ocr CLI for file selection and review rules while the host agent does the code review itself, file by file, with required coverage reporting.↻Review every changed file, report coverage.Xcode Project Analyzerby AvdLee · 4K installsThe Xcode Project Analyzer skill (xcode-project-analyzer) audits Xcode project configuration, build settings, scheme order and script phases to find build-time wins, recommendation-first with approval gates.↻Project-level build waste found and ranked.
Understand Diffby egonex-ai · 4K installsUnderstand diff analyzes git diffs and pull requests against a codebase knowledge graph to show changed components, affected neighbours, touched layers and a risk rating.↻Every PR gets an impact map.
Discriminated Unionsby mkosir · 4K installsThe typescript-discriminated-unions skill applies, reviews and explains discriminated unions in TypeScript to model mutually exclusive variants, prevent invalid states and enforce exhaustiveness.↻Invalid states removed from the code.
Auto Create PRby open-mercato · 4K installsAuto Create PR (the om-auto-create-pr skill) runs a free-form task end to end: a Progress-tracked plan, phase-by-phase commits in an isolated worktree, a validation gate, a review pass and a labelled PR.↻Task brief shipped as a PR
Vibe Securityby raroque · 4K installsVibe Security audits codebases for the security holes AI assistants commonly introduce: exposed secrets, broken access control, missing auth, client-trust and payment flaws.↻Common AI-coded security holes caught.
Playwright E2E Testingby alinaqi · 3K installsThe playwright e2e testing skill sets up end-to-end tests with Playwright: page objects, role-based locators, shared auth, network mocks, visual checks, dead link detection and CI.↻No broken flows or dead links.
Three.js Gameplay Systemsby majidmanzarpour · 3K installsThe three.js gameplay systems skill builds playable browser game loops: a Vite and TypeScript starter, design briefs, core loops, levels, input, camera, physics, scoring and game feel.↻A game loop you can play.
R8 Analyzerby android · 3K installsR8 Analyzer analyses Android build files and R8 keep rules to find redundancies, overly broad rules and rules that subsume library consumer rules, reporting them to trim app size.↻Leaner, non-redundant R8 keep rules.
Arize Datasetby arize-ai · 3K installsThe arize dataset skill creates, manages and queries Arize datasets with the ax CLI: create from files, append examples, export versions and verify completeness.↻A growing, verified evaluation dataset.
Paperclip create pluginby paperclipai · 3K installsThe paperclip create plugin skill scaffolds, installs and verifies external Paperclip plugins with the paperclipai CLI, for developers extending a local Paperclip instance.↻Plugin builds, tests and installs.Arize Evaluatorby arize-ai · 3K installsThe arize evaluator skill creates LLM-as-judge and code evaluators on Arize with the ax CLI, wires column mappings, and runs evaluation tasks on spans or experiments.↻Score new spans every week.Three.js Debug Profilerby majidmanzarpour · 3K installsThree.js Debug Profiler debugs and profiles Three.js browser games: blank canvases, runtime and loading bugs, input issues, and draw call, texture, memory and shader costs.↻Root causes fixed, bottlenecks measured.
Netlify Edge Functionsby netlify · 3K installsThe netlify edge functions skill helps an agent write, route, cache and debug Deno-based edge handlers on Netlify, from auth gates and geo redirects to streaming SSR.↻No silent edge functions ship.Netlify Formsby netlify · 3K installsThe netlify forms skill helps an agent add and debug forms on Netlify sites: detection at deploy time, AJAX submission, file uploads, spam filtering and notifications, including forms in Next.js, Nuxt, SvelteKit, Astro and Gatsby.↻Every form still gets submissions.
Tigris SDK Guideby tigrisdata · 3K installsThe tigris sdk guide skill tells the agent which SDK to use with Tigris per language: native SDKs for TypeScript and Go, a boto3 extension for Python, AWS SDKs only for Ruby and PHP, and the tigris CLI instead of aws s3.↻Code uses native Tigris SDKs.Convex Realtimeby waynesutton · 3K installsThe convex realtime skill helps build reactive UIs on Convex subscriptions with useQuery patterns, optimistic updates, live pagination, presence and less subscription churn.↻Live queries stay small and indexed.React Vite Best Practicesby asyrafhussin · 3K installsThe react vite best practices skill gives React developers 23 rules for Vite builds, code splitting, HMR, assets, environment variables and bundle analysis, ranked by impact.↻Your Vite bundle stays lean.
Code Polishby paulrberg · 3K installsThe code polish skill simplifies and reviews changed code over a fixed scope, fixing evidence-backed defects by risk with language and security profiles, then verifies with lint, types and tests.↻Changed code ends up simpler.
Redis Connectionsby redis · 3K installsThe redis connections skill gives developers client-side guidance for Redis: connection pools or multiplexing, pipelining, SCAN instead of KEYS, RESP3 client-side caching and sensible connect and read timeouts.↻Redis clients stay pooled and fast.
Rsbuild Best Practicesby rstackjs · 3K installsThe rsbuild best practices skill helps developers configure, review and troubleshoot Rsbuild apps: config layout, CLI commands, type checking, bundle size, assets, source map safety, debugging and profiling.↻Rsbuild builds stay lean and safe.
Slack Web APIby slackapi · 3K installsThe slack web api skill helps developers find the right Slack method, read its scopes, token type and arguments from the live docs, call it safely, page with cursors and handle rate limits and errors.↻Slack API calls stay correct.
Architecture Compassby stark-ai-de · 3K installsThe architecture compass skill sets up ADR governance, audits architecture and drift, and plans or runs bounded refactors for teams that want binding, traceable architecture decisions in their repository.↻Code keeps to your ADRs.CodeGraph and ast-grepby stark-ai-de · 3K installsThe codegraph and ast-grep skill sets up, updates or diagnoses CodeGraph and the ast-grep CLI so coding agents get semantic repository scope and structural code search, with persisted guidance and safe migrations.↻Code analysis tools stay ready.
Capacitor Best Practicesby cap-go · 3K installsThe capacitor best practices skill reviews and sets up Capacitor apps: version alignment, capacitor.config audit, plugin patterns, native project checks and release prep, with fixes for common errors.↻Capacitor app stays release ready.Propagateby juxt · 3K installsThe propagate skill generates tests from Allium specifications: assertion, property-based and state machine tests mapped to real code, with every obligation reconciled, for teams that keep a spec as the source of truth.↻Spec changes regenerate the tests.Weedby juxt · 3K installsThe weed skill compares Allium specifications with implementation code, finds where they diverged, classifies each divergence and helps resolve it, for teams that keep specs and code in sync.↻Spec and code stay aligned.
jscpdby kucherenko · 3K installsThe jscpd skill runs the jscpd copy-paste detector across 220+ languages, finds exact, renamed and near-miss clones, and ranks refactoring hotspots with a codebase summary.↻Copy-pasted code stops growing.
Web Debugby sentimony · 3K installsThe web debug skill lets agents test local web apps with Python Playwright scripts: start servers, wait for rendering, collect console and network signals, take screenshots and tell real errors from noise.↻Console errors caught before users.
Effectby kitlangton · 2K installsThe effect skill helps TypeScript developers build production apps with Effect v4: services, layers, schemas, config, schedules, caches, streams, HTTP clients and tests.↻Effect code follows v4 defaults.
Qodo Get Rulesby qodo-ai · 2K installsThe qodo get rules skill helps development teams retrieve their organization's Qodo coding rules for a task and apply them by severity while the agent writes or reviews code.↻Team coding rules applied every task.
Hugging Face Gradioby huggingface · 2K installsThe hugging face gradio skill helps Python developers build Gradio web UIs, demos and chatbots with Interface, Blocks and ChatInterface, and call running apps with the gradio CLI.↻Your Gradio endpoints keep answering.
OpenAI Agents SDKby laguagu · 2K installsThe openai agents sdk skill helps Python developers build agents with the openai-agents package: tools, handoffs, guardrails, sessions, streaming, tracing and offline testing.↻Agent changes verified offline first.
Frontend Testing Best Practicesby sergiodxa · 2K installsThe frontend testing best practices skill helps frontend developers write and review tests that favor E2E over unit tests, keep mocking minimal and test behavior instead of implementation.↻Tests check behavior, not implementation.
Botpress ADKby botpress · 2K installsThe botpress adk skill helps developers build bots with the Botpress Agent Development Kit: actions, tools, workflows, conversations, tables, knowledge bases, triggers, integrations and Zai.↻ADK bot checked before every session.
Astro Frameworkby delineas · 2K installsThe astro framework skill helps web developers build fast content sites with Astro islands: hydration directives, server islands, Content Layer loaders, SSR adapters, sessions, actions and i18n.↻Astro pages ship only needed JavaScript.Pydanticby pydantic · 2K installsThe pydantic skill helps Python developers model data with Pydantic: field constraints, validators, coercion, forward annotations and safe model hierarchies with discriminated unions.↻Pydantic models use built-in validation.
RTK TDDby rtk-ai · 2K installsThe rtk tdd skill enforces red-green-refactor test-driven development for Rust, with cfg(test) modules, Arrange-Act-Assert patterns and a fmt, clippy and test gate before commit.↻Every Rust change starts red.
Test Harnessby zernie · 2K installsThe test harness skill helps Claude Code users test hooks, skills, agents, settings and CLAUDE.md with vigiles, picking unit, deterministic or eval tiers and keeping eval results honest in CI.↻Hooks and skills always tested.
Appwrite TypeScriptby appwrite · 2K installsThe appwrite typescript skill helps web, React Native and Node.js developers use the Appwrite SDK for auth, TablesDB queries, storage, realtime, functions and SSR sessions.↻Appwrite code follows current SDK patterns.
bktby avivsinai · 2K installsThe bkt skill helps developers operate Bitbucket Cloud or Data Center from the terminal: repositories, pull requests, branches, issues, pipelines, permissions and webhooks.↻Failing Bitbucket PRs flagged daily.Appwrite CLIby appwrite · 2K installsAppwrite cli is a skill for developers who use the Appwrite CLI well: config vs client, pull and push, type-safe code with appwrite generate, query flags, CI auth, regional endpoints and function variables.↻Schema, config and types stay synced.
Node.js Coreby mcollina · 2K installsNode.js core is a skill for people who contribute to or debug nodejs/node: commit and PR style, the build-lint-test cycle, native addon crashes, V8 deoptimizations, node-gyp builds and libuv issues.↻Commits that pass CI first.
Playwright Automationby petrkindlmann · 2K installsPlaywright automation is a skill for QA and frontend engineers who write production-grade Playwright tests in TypeScript with page objects, fixtures, user-facing locators, parallel runs, CI sharding and version checks.↻E2E tests that stay stable.
Code Quality Guideby tursodatabase · 2K installsThe Code Quality Guide skill gives correctness rules and Rust patterns for a production database: assert invariants, crash over corrupt, exhaustive matches, no over-engineering and careful index ordering.↻Changes that crash, never corrupt.Flutter Architectureby madteacher · 2K installsFlutter architecture is a skill for Flutter developers who design, review or refactor app structure with MVVM, layered boundaries, feature-first or layer-first layout, repositories, services and Command and Result patterns.↻Flutter features keep clean layer boundaries.
CMakeby mohitmishra786 · 2K installsCmake is a skill for C and C++ developers who write modern target-first CMakeLists.txt files: out-of-source builds, generators, dependencies with find_package and FetchContent, sanitizers, cross-compilation and presets.↻Green C++ builds in every config.OWASP Security Checkby sergiodxa · 2K installsOwasp security check is a skill for web developers that audits web apps and REST APIs against 20 rules based on the OWASP Top 10, from access control and injection to headers, CORS and rate limiting.↻Security flaws caught before deploy.
Design Control Loopby humanlayer · 2K installsDesign control loop is a skill for engineering teams that interviews you to design an agentic control loop for your codebase, with sensor, controller and actuator, then builds it as local scripts plus a scheduled agent workflow.↻A scheduled loop driving your codebase.
React Routerby remix-run · 2K installsReact router helps build apps in framework, data, declarative and unstable RSC modes: routes, loaders, actions, forms, fetchers, navigation, pending UI, SSR and upgrades, using the installed version's docs.↻Route changes match the app's mode.
tsdownby rolldown · 2K installsTsdown is a skill for library authors that bundles TypeScript and JavaScript libraries with tsdown and Rolldown: declarations, ESM and CJS formats, dependency handling, package validation and migration from tsup.↻Library builds valid before publish.
AI Slop Cleanerby yeachan-heo · 2K installsAI Slop Cleaner cleans AI-generated code with a regression-safe, deletion-first workflow: lock behavior, plan, then remove dead code, duplication and needless abstraction, with an optional reviewer-only mode.↻Bloated AI code trimmed, behavior kept.
Mod Any Gameby rehan-remade · 1K installsMod any game is a skill for PC gamers and modders that takes a mod idea for a game they own to a working, recorded result: recon, route choice, safe lab, vertical slice, generated assets and in-game checks.↻Mods working in the real game.RTK TDD Rustby rtk-ai · 1K installsThe RTK TDD Rust skill enforces red-green-refactor for RTK filter development: real command fixtures, a 20% token savings test, insta snapshots, edge case tests and a fmt, clippy and test gate.↻Every new filter proves its savings.
Lightweight Chartsby tradingview · 1K installsThe lightweight charts skill helps developers build TradingView lightweight-charts v5 charts: series, scales, realtime data, panes, markers, plugins, SSR loading and React or Vue wrappers.↻Chart code stays on v5.Extract Errorsby react · 1K installsThe extract errors skill runs the React error code extraction, reports new error messages that need codes and checks that codes are up to date.↻Every new error gets a code.Issue Triageby rtk-ai · 1K installsThe issue triage skill audits open GitHub issues with gh, categorizes them, finds likely duplicates, links PRs, rates risk and staleness, then drafts comments, labels and closes that it runs only after you approve.↻Sort open issues, act after approval.Debuggingby tursodatabase · 1K installsShows how to hunt bugs in the Turso database.↻Each odd result traced to its cause and logged. Bugs found faster.Testingby tursodatabase · 1K installsThe testing skill from the Turso repo tells agents which test type to use, how to run them and how to write .sqltest, TCL and Rust integration tests, including converting TCL tests to .sqltest.↻Legacy TCL tests move to sqltest.React feature flagsby react · 1K installsThe react feature flags skill explains React's flag files, @gate pragmas, inline gate() checks and __VARIANT__ flags, and how to debug channel-specific test failures.↻Flags consistent across every fork.React testby react · 1K installsThe react test skill runs tests for the React codebase on the source, experimental, stable, classic and www channels, with the right yarn command and flags.↻React tests run per channel.Flowby react · 1K installsThe flow skill runs Flow type checking in the React repository for a chosen renderer and reports type errors with file locations, for contributors who change React code or see Flow errors.↻No Flow errors reach main.
Generate Translationsby payloadcms · 569 installsThe generate translations skill helps Payload contributors add English keys to core or plugin packages and run the pnpm scripts that translate them into every supported language.↻Every new key gets translated.
Mem0 Integrateby mem0ai · 397 installsThe mem0 integrate skill adds Mem0 memory to an existing Python or TypeScript backend with a test-first pipeline, approval gates and a feature flag.↻Mem0 added behind a feature flag.Mem0 Test Integrationby mem0ai · 322 installsMem0 Test Integration verifies a Mem0 integration on its branch: install, static checks, the repo's tests with the flag off and on, a real smoke call and an end-to-end flow, then a scorecard.↻A Mem0 integration proven to run.Studio Testingby supabase · 321 installsThe studio testing skill gives the testing strategy for Supabase Studio: extract logic into .utils.ts files, test every permutation, use component tests only for complex UI and E2E tests for shared features.↻Every Studio change gets right tests.Triage CI Flakeby payloadcms · 301 installsThe triage ci flake skill gives agents a reproduce-first routine for e2e failures in CI: aggregate failing checks, rerun the exact Playwright test with dev and prod builds, fix and confirm with repeated runs.↻Flaky e2e tests reproduced, then fixed.Audit Dependenciesby payloadcms · 127 installsThe audit dependencies skill helps pnpm monorepo maintainers fix vulnerabilities from pnpm audit the simplest way: direct bump, lockfile update, or a justified override, then a PR with advisory links.↻Vulnerable dependencies fixed the simplest way.Fix Bugby remix-run · 102 installsThe fix bug skill takes a React Router GitHub issue and works through it: reads the issue, checks the reproduction, writes a failing test, makes a minimal fix, runs tests, lint and typecheck, and prepares a PR.↻Fix reported bugs with tests.Finish Lineby remix-run · 99 installsThe finish line skill brings a blocked React Router community pull request to a mergeable state by adding missing tests, change files or docs on the contributor branch, with maintainer approval before any push.↻Blocked community PRs get merged.UI4 Convert Testsby payloadcms · 66 installsThe ui4 convert tests skill helps Payload developers find and fix e2e tests broken by UI changes, by analyzing the diff for selector, structure and text changes and updating helpers first.↻E2E tests follow UI changes.
Diagnosing CI and Merge Bottlenecksby PostHog · 2 installsThe diagnosing ci and merge bottlenecks skill helps engineering teams use PostHog engineering analytics tools to find slow or flaky workflows, time to merge and stuck pull requests.↻CI long pole named weekly.Verification Before Completionby Jesse Vincent · 249K installsThe verification before completion skill makes your agent prove a task is done with evidence before it says so.No habit yetAgent Skillsby Addy Osmani · 109K starsThe agent skills from Addy Osmani give your agent production engineering habits: performance, testing and code that holds up.No habit yetClaude Code skills, answered.
What are the best Claude code skills?
The list above starts with the ones people install most that already have a habit. Each card says what the skill does and what its habit does by itself.
What is the difference between a skill and a habit?
A skill works once, when you ask. A habit runs the skill by itself: it starts on a schedule or event, checks its own work and keeps notes.
How do I use these code skills?
Paste the setup prompt above into Claude Code, Codex or Cursor. Your agent then finds the right skill for each task.