GitHub - spencerpauly/awesome-cursor-skills: A curated list of awesome skills for Cursor

12 min read Original article ↗

A curated list of awesome skills for Cursor, the AI code editor.

Skills are reusable SKILL.md instruction files that teach the AI agent how to perform specific tasks — from setting up analytics to scaffolding entire projects. In cursor, they live in .cursor/skills/ (personal) or project directories and are automatically discovered by the agent.

Contents

Skills

Ready-to-use SKILL.md files you can copy into your .cursor/skills/ directory. Each one teaches the agent a specific engineering workflow.

Cursor-Native

Skills that harness Cursor's unique agent capabilities — things only an AI inside Cursor can do.

  • suggesting-cursor-rules - If you keep correcting the agent on the same convention, it suggests a .cursor/rules/ file to encode it permanently.
  • suggesting-cursor-hooks - If you keep asking the agent to run the same check (lint, types, tests), it suggests a .cursor/hooks.json hook to automate it.
  • switching-projects - Switch the active workspace to a different project using the cursor-app-control MCP, without opening a new window.
  • saving-workspace-context - Automatically persist research, decisions, and learnings to workspace files so knowledge survives across conversations.
  • visual-qa-testing - Open the app in Cursor's built-in browser, take screenshots, check console errors, and audit network requests after making changes.
  • verifying-in-browser - Start the dev server, open the app side-by-side with your code, and verify rendering, console, and network health.
  • profiling-performance - Profile CPU performance of a running app using Cursor's browser profiler to capture call stacks and identify slow functions.
  • screenshotting-changelog - Generate visual before/after PR descriptions by screenshotting UI changes across branches.
  • best-of-n-solving - Try multiple approaches to a hard problem in parallel using isolated git worktrees, then pick the best solution.
  • parallel-exploring - Explore a large codebase fast by launching multiple read-only subagents that investigate different areas simultaneously.
  • grinding-until-pass - Keep iterating autonomously — fix, run, check, repeat — until tests pass, the build succeeds, or linting is clean.
  • finding-dev-server-url - Scan running terminals for dev server URLs and ports, report what's running where, and open the app in Cursor's browser.
  • monitoring-terminal-errors - Watch a running process for crashes and stack traces, navigate to the failing file, and fix it automatically.
  • detecting-port-conflicts - Detect EADDRINUSE errors, find what's using the port, and resolve by killing the process or suggesting an alternative.
  • tailing-build-output - Monitor a build process for warnings and errors as they stream, fix issues before the build finishes.
  • responsive-testing - Open the app at multiple viewport sizes (mobile, tablet, desktop), screenshot each, and report layout breakage.
  • dark-mode-testing - Toggle between light and dark mode in the browser, screenshot both, and flag missing token mappings or contrast issues.
  • accessibility-auditing - Use the browser's aria tree to audit for missing labels, broken tab order, ARIA misuse, and contrast issues.
  • form-testing - Fill and submit every form with valid/invalid data using the browser, verifying validation, error states, and success flows.
  • parallel-test-fixing - When multiple tests fail, assign each to a separate subagent that fixes it independently in parallel.
  • codebase-onboarding - Launch parallel explore subagents to investigate architecture, data models, auth, APIs, and deployment — then synthesize an onboarding doc.
  • comparing-branches-visually - Run two branches on different ports, screenshot the same pages from each, and produce a visual diff for PRs.
  • auto-type-checking - Run tsc --noEmit after file edits to catch type errors immediately. Supports Cursor hooks for automation.
  • suggesting-skills - When you're struggling with a task a known skill handles, the agent suggests installing it.
  • parallel-ci-triage - Fetch failing GitHub Actions logs, assign each failing job to a parallel subagent, fix independently, then merge and re-run CI.
  • parallel-code-review - Run four read-only subagents in parallel — security, performance, correctness, readability — and merge into one review report.
  • network-request-auditing - Use the browser's network log to flag failed requests, slowness, duplicates, and risky payloads after real user interactions.
  • recording-browser-flow-as-test - Walk a flow in Cursor's browser, log each step, then emit a Playwright spec using role/name locators from the accessibility tree.
  • building-skills-from-patterns - Turn repeated multi-step workflows into a new SKILL.md under .cursor/skills/ so the agent can reuse them.

Analytics & Tracking

  • adding-analytics - Add PostHog event tracking, page views, feature flags, and session replay to any web app.
  • adding-feature-flags - Add feature flags for gradual rollouts and A/B testing using PostHog or a simple local implementation.
  • posthog-llm-analytics - Instrument LLM calls with token usage, latency, cost tracking, and model comparison.
  • posthog-migrations - Migrate from other analytics providers (Amplitude, Mixpanel, GA) to PostHog.

Error Tracking & Monitoring

Authentication & Payments

  • adding-auth - Add OAuth login, session management, and protected routes with Auth.js (NextAuth).
  • adding-stripe - Integrate Stripe checkout, subscriptions, webhooks, and customer portal.

Testing

  • adding-e2e-tests - Set up Playwright with config, example tests, page objects, and CI integration.
  • writing-tests - Analyze existing code and write comprehensive unit and integration tests with proper mocking, edge cases, and assertions.
  • python-tdd-with-uv - Test-driven development in Python using uv — red-green-refactor cycle with vertical slicing and fast dependency management.
  • mattpocock-tdd - Vertical-slice TDD for AI agents — one test, one implementation, repeat. Prevents over-engineering and speculative tests.
  • anthropic-webapp-testing - Automated browser testing for web apps with screenshot verification and interaction flows.
  • api-smoke-testing - Discover API routes from the codebase, hit every endpoint, and report which ones return errors.

Workflow

  • babysitting-pr - Monitor an open PR for CI failures, review comments, and merge conflicts — then fix them automatically to keep the PR merge-ready.
  • creating-pr - Create clean, review-ready pull requests with conventional titles, structured descriptions, and linked issues.
  • writing-commit-messages - Write conventional commit messages with type prefixes, scopes, and meaningful descriptions.
  • incident-response - Handle production incidents — triage severity, mitigate, communicate status, and write blameless postmortems.
  • systematic-debugging - Structured debugging methodology — reproduce, isolate, hypothesize, and verify using git bisect, binary search, and logging.
  • chatcrystal - Local-first memory recall and writeback skills for AI coding sessions via ChatCrystal Core and MCP.

Infrastructure & DevOps

  • adding-docker - Dockerize any app with a multi-stage Dockerfile, docker-compose, and .dockerignore.
  • setting-up-ci - Set up a GitHub Actions CI/CD pipeline with linting, testing, type-checking, and deployment.
  • setting-up-terraform - Infrastructure-as-code with provider config, modules, remote state, and CI integration.
  • antonbabenko-terraform - Terraform and OpenTofu skill — testing, modules, CI/CD, and production patterns.
  • kubernetes-deploying - Deploy to Kubernetes — Deployments, Services, Ingress, ConfigMaps, health checks, and autoscaling.

Code Quality & Security

  • reviewing-code - Thorough code review focused on correctness, maintainability, performance, and best practices.
  • auditing-security - Systematic security audit checking for OWASP Top 10 vulnerabilities, secrets exposure, and insecure patterns.
  • auditing-performance - Audit bundle size, rendering, database queries, and Core Web Vitals.
  • sentry-code-simplifier - Refactor for clarity, consistency, and maintainability — eliminates dead code, fixes naming, and reduces complexity.
  • sentry-find-bugs - Scan local branch changes for bugs, security vulnerabilities, and code quality issues.
  • sentry-code-review - Code review following Sentry's engineering practices.
  • sentry-security-review - Security code review for injection, XSS, auth bypass, and IDOR vulnerabilities.
  • sentry-django-perf-review - Django-specific performance review — N+1 queries, select_related, caching, and serialization.
  • sentry-skill-scanner - Scan agent skills for security issues — prompt injection, exfiltration, and unsafe tool use.
  • verifying-markdown-formatting - Verify headings, lists, links, code blocks, spacing, and style consistency in Markdown files.
  • fixing-broken-links - Crawl all URLs in a file, test each for HTTP 200, and fix or replace any broken links.

Dependencies

  • updating-npm-package - Safely update an npm package: check npmjs.com for the latest version, read release notes, auto-apply minor updates, and for major updates find the migration guide and produce a detailed validation report.

Frontend & UI

  • using-ui-stack - Enforce a design system (8px grid, color tokens, typography, dark mode, 5-state interactions) on every AI-generated component.
  • converting-css-to-tailwind - Convert plain CSS stylesheets to Tailwind utility classes — selectors, media queries, pseudo-classes, animations, and arbitrary values.
  • converting-css-modules-to-tailwind - Migrate CSS Modules to Tailwind — handles styles.xxx removal, composes, conditional classNames, SCSS features, and cleanup.
  • anthropic-frontend-design - Generate polished, production-ready frontend UI with consistent styling and responsive layouts.
  • shadcn-ui - Managing shadcn components — adding, searching, debugging, styling, and composing UI.
  • vercel-react-best-practices - 40+ rules for React/Next.js performance including eliminating request waterfalls and bundle optimization.
  • vercel-web-design-guidelines - UI code auditing for accessibility, UX, and performance compliance.
  • vercel-react-view-transitions - Implement the View Transitions API in React/Next.js for smooth page and component animations.
  • vercel-composition-patterns - Component composition, code splitting, and server/client boundary patterns for Next.js.
  • react-native-patterns - Build mobile apps with React Native and Expo — navigation, platform-specific code, performance, and native modules.

Planning & Architecture

Documentation

  • adding-api-docs - Generate OpenAPI/Swagger documentation with interactive docs UI.
  • anthropic-doc-coauthoring - Structured workflow for co-authoring technical documentation with an AI agent.
  • anthropic-docx - Generate Word documents (DOCX) programmatically with formatting, tables, and images.
  • anthropic-pdf - Generate PDF documents with layouts, charts, and structured content.
  • anthropic-pptx - Create PowerPoint presentations with slides, layouts, and visual content.
  • anthropic-xlsx - Build Excel spreadsheets with formulas, charts, and data formatting.

Utilities

  • exporting-to-png - Export code snippets, diagrams, terminal output, or UI components to PNG images via headless browser or CLI tools.
  • generating-images - Generate or edit images (icons, logos, blog heroes, OG images, illustrations, mockups) using OpenAI gpt-image-2. Supports text-to-image, image-to-image, masked inpainting, and parallel batch jobs.
  • prompt-engineering - Write effective LLM prompts — system prompts, few-shot examples, chain-of-thought, and structured output.
  • seo-auditing - Audit technical SEO — meta tags, structured data, Open Graph, sitemaps, and Core Web Vitals.
  • seo-analysis - Full SEO audit using Search Console, URL inspection, PageSpeed, technical crawling, metadata checks, schema review, and a prioritized 30-day action plan.
  • writing-copy - Write marketing copy for landing pages, CTAs, emails, microcopy, and product descriptions.
  • concise - Chinese-first concise mode skill. Compresses Cursor agent replies on two layers (expression + content) with auto-relax for safety, multi-step, and parameter-heavy cases. Works across Cursor, Claude Code, and Codex CLI.

Plugins

Official Cursor marketplace plugins with bundled skills. Install via Cursor Settings > Plugins.

  • Figma - (generate-design, code-connect-components + 5 more) Design-to-code and design system management.
  • Linear - Issues, projects, documents, and sprints via MCP.
  • Slack - Channel search, messaging, and Slack actions via MCP.
  • Datadog - (datadog-mcp-setup) Query logs, metrics, traces, and dashboards.
  • Stripe - (stripe-best-practices, upgrade-stripe) Payment integration and SDK migration.
  • Firebase - (firebase-ai-logic, firebase-auth-basics + 9 more) Backend, auth, and AI infrastructure.
  • Shopify - (shopify-admin, shopify-custom-data + 14 more) GraphQL, Liquid, and extensions.
  • dbt Labs - (adding-dbt-unit-test, building-dbt-semantic-layer + 7 more) Data modeling and analytics engineering.
  • Sentry - (sentry-code-review, sentry-browser-sdk + 26 more) Error tracking and SDK integration.
  • Vercel - (ai-architect, deployment-expert, performance-optimizer + 25 more) Deploy and optimize web apps.
  • Svelte - (svelte-file-editor, svelte-code-writer + 2 more) Svelte 5 development and validation.
  • Elastic - (cloud-create-project, cloud-manage-project + 31 more) Elasticsearch and Observability.
  • Postman - (api-readiness-analyzer, postman-routing + 11 more) API lifecycle management.
  • Sanity - (content-modeling-best-practices, seo-aeo-best-practices + 6 more) CMS and content workflows.
  • Langfuse - (langfuse) LLM tracing, prompt management, and evaluation.
  • CockroachDB - (cockroachdb-sql-patterns, cockroachdb-app-patterns + 2 more) Distributed DB management.
  • Encore - (add-infrastructure, create-service + 20 more) TypeScript/Go backends with auto infrastructure.
  • AWS Amplify - (amplify-workflow) Full-stack apps with auth, data, and storage.
  • AWS Serverless - (aws-lambda, api-gateway + 2 more) Serverless application lifecycle.
  • Pendo - (account-health, feature-adoption + 2 more) Product analytics and session replays.
  • GitLab - (gitlab-ci-author, backlog-health + 7 more) Issues, MRs, and CI/CD pipelines.
  • Sourcegraph - (searching-sourcegraph, sourcegraph-deepsearch) Code search across repos.
  • Miro - (miro-mcp, diagram + 4 more) Board context, diagrams, and code generation.
  • Cloudinary - (cloudinary-docs, cloudinary-transformations) Media management and optimization.
  • Appwrite - (appwrite-cli, appwrite-web + 8 more) BaaS with database, auth, and storage.
  • ClickUp - Task management, time tracking, and collaboration via MCP.
  • Box - (box) Content management, AI Q&A, and extraction.
  • Chargebee - (chargebee-integration) Billing operations and webhook handling.
  • Circle - (bridge-stablecoin, use-circle-wallets + 7 more) Stablecoin payments and smart contracts.
  • Runlayer - (mcp-builder, mcp-security-audit, plugin-builder) MCP security and governance.
  • Superpowers - (code-reviewer, brainstorming + 13 more) TDD, debugging, and collaboration.
  • Omni - (omni-content-builder, omni-ai-optimizer + 9 more) Analytics and dashboard creation.
  • Braintrust - AI evaluation, experiments, and evaluation logs via MCP.
  • Parallel - (parallel-web-search, parallel-deep-research + 2 more) Web search and data enrichment.
  • Tavily - (tavily-cli, tavily-extract + 4 more) Web search, crawling, and deep research.
  • Astronomer - (airflow, airflow-plugins + 19 more) Data engineering and Airflow integration.
  • Meta Reality Labs - (hz-immersive-designer, hz-new-project-creation + 11 more) Quest and Horizon OS development.
  • Plain - Support threads, customer management, and help center articles.
  • Turbopuffer - Vector and full-text search database integration.
  • Superpower Builder - Interview-driven meta-builder that turns recurring tasks into reusable SKILL.md files, routed by kind (workflow, discipline, content, subagent) with baseline-vs-with-skill pressure-testing on each generated skill. Ships 25 bundled superpowers covering dev (TDD, plan/execute) and non-dev (research, writing, decisions) work. Installs via Cursor plugin descriptor (.cursor-plugin/plugin.json).

Cursor Resources

Cursor Rules

Cursor Rules (.cursorrules / .cursor/rules/) are complementary to skills — rules are always-on conventions, skills are on-demand workflows.

Learning

Directories

Tools

Contributing

Contributions welcome! Please read the contribution guidelines first.

If you know of a great Cursor skill that isn't listed here, please open a PR. Include:

  • A link to the skill or repository
  • A brief description of what it does
  • Which category it belongs to