AgentSkills.site

Skills library

37 real skills, evaluated — not just listed.

A curated set of agent skills with real source repositories, checked directly against their own docs and manifests. Each entry explains what a skill actually does, when it earns a place in your setup, when it doesn’t, what it depends on, and how it compares to the skill next to it — not a mirror of its README.

Not sure where to start from a goal instead of a browse? Ask Doctor — it recommends a small stack from this same set and links straight back into the profiles below.

Capability

Agent

Provenance

37 of 37 skills

Migrating from another agent

  1. 01migrate-to-codexMigrates instruction files, skills, agents, and MCP config from another agent into Codex's own files, without touching the source agent's files.CodexOfficial OpenAI

Planning and orchestration

  1. 01orchestrateFans a large task out across parallel Cursor cloud agents via the Cursor SDK — planners publish tasks, workers hand off, a script reconciles the tree from disk and git.CursorOfficial Cursor

Writing your own skills

  1. 01example-skillsAnthropic's reference bundle: skill-creator, mcp-builder, webapp-testing, frontend-design, canvas-design, theme-factory, and more — twelve skills installed together as one plugin.Claude CodeOfficial Anthropic
  2. 02skill-creatorCodex's own bundled tool for authoring a new skill, invoked with $skill-creator.CodexOfficial OpenAI
  3. 03/create-skillCursor's built-in skill for writing a new skill, the recommended starting point over hand-authoring frontmatter.CursorOfficial Cursor
  4. 04create-pluginMeta-workflow for scaffolding a marketplace-ready Cursor plugin — manifest, skills, rules, agents, and commands together — plus a pre-submission quality check.CursorOfficial Cursor

Development discipline

  1. 01superpowersA structured development methodology bundling brainstorming, planning, TDD, systematic debugging, and code review into one workflow rather than one narrow task.Claude CodeCommunity obra (Jesse Vincent)

Tool and API integration

  1. 01claude-apiClaude API and SDK documentation, packaged so Claude can reference method signatures directly while you build an LLM-powered app.Claude CodeOfficial Anthropic
  2. 02nvidia-skillsNVIDIA's own router across GPU acceleration, CUDA, agent training, inference, robotics, Physical AI, and simulation workflows.Claude CodeCodexCursor— portableVendor NVIDIA
  3. 03n8n Skillsn8n's own bundle pairing 13 workflow-lifecycle skills with its instance-level MCP server, so an agent builds and edits n8n workflows correctly the first time.Claude CodeCodexVendor n8n
  4. 04cli-for-agentDesign and review guidance for building CLIs meant to be driven by coding agents — non-interactive flags, layered --help, stdin/pipelines, fast errors, idempotency, --dry-run.CursorOfficial Cursor
  5. 05cursor-sdkIntegration guidance for building on the Cursor TypeScript SDK (@cursor/sdk) — invocation patterns, runtime choice, auth, streaming, MCP, and error handling.CursorOfficial Cursor

Security review

  1. 01audit-context-buildingReads a codebase function by function, records what each one assumes and depends on, and writes the results to files instead of filling up the conversation — built for pre-audit codebase understanding.Claude CodeCursor— portableVendor Trail of Bits
  2. 02security-threat-modelProduces a repository-grounded threat model — trust boundaries, assets, attacker capabilities, abuse paths, mitigations — written to a markdown file. Roughly half its description is negative triggers.CodexOfficial OpenAI
  3. 03thermosRuns two parallel Cursor subagents — a deep correctness/security audit and a strict maintainability rubric — and synthesizes both into one prioritized branch review.CursorOfficial Cursor

Code review

  1. 01review-agentCodex's bundled skill for reviewing changes — the current in-box option, no install needed.CodexOfficial OpenAI
  2. 02Cursor's built-in review skills/review, /review-bugbot, and /review-security ship with Cursor by default — general review, automated bug-hunting review, and a security-focused review, with nothing to install.CursorOfficial Cursor
  3. 03parallel-code-reviewRuns four read-only subagents in parallel — security, performance, correctness, readability — and merges the results into one prioritized review.CursorCommunity spencerpauly
  4. 04Cursor Team KitA 17-skill Cursor plugin for CI, code review, shipping, and test reliability — loop-on-ci, review-and-ship, fix-ci, get-pr-comments, weekly-review, and more, installed together.CursorOfficial Cursor
  5. 05pr-review-canvasRenders a pull request diff as an interactive Cursor Canvas organized by reviewer value — core logic first, wiring condensed, boilerplate summarized.CursorOfficial Cursor

GitHub PR workflow

  1. 01gh-address-commentsFinds the open PR for your current branch and works through its review comments with the gh CLI, checking gh auth status first rather than failing halfway through.CodexOfficial OpenAI

Turning material into a skill

  1. 01book-to-skillConverts a PDF, EPUB, DOCX, or plain-text document into a structured skill, extracting frameworks and mental models rather than producing a summary.Claude CodeCommunity virgiliojr94

Documentation and knowledge capture

  1. 01continual-learningAutomatically and incrementally keeps AGENTS.md up to date, mining new session transcripts on a cadence and updating matching bullets in place rather than rewriting the file.CursorOfficial Cursor
  2. 02docs-canvasRenders architecture notes, API references, runbooks, or codebase walkthroughs as a navigable Cursor Canvas instead of a flat markdown file.CursorOfficial Cursor

Database and backend

  1. 01supabaseSupabase's official, comprehensive skill covering Database, Auth, Edge Functions, Realtime, Storage, Vectors, Cron, and Queues, plus client-library and SSR integration.Claude CodeCursorVendor Supabase
  2. 02supabase-postgres-best-practicesPostgres performance-optimization guidelines from Supabase, covering query performance, connection management, schema design, RLS, and monitoring — prioritized by impact.Claude CodeCursorVendor Supabase
  3. 03databases-on-awsDatabase guidance for the AWS portfolio — schema design, queries, migrations, and multi-tenant patterns — currently available for Aurora DSQL.Claude CodeCodexVendor AWS Labs

Document generation

  1. 01document-skillsAnthropic's official bundle for generating and editing Word, PDF, PowerPoint, and Excel files, with descriptions tuned to trigger on phrases like "the deck" or "Word doc."Claude CodeOfficial Anthropic
  2. 02pdfReads, creates, and reviews PDFs where layout matters, rendering pages for visual checks rather than only extracting text.CodexOfficial OpenAI

Testing and QA

  1. 01playwrightDrives a real browser from the terminal for navigation, form filling, screenshots, and data extraction — CLI-first, and won't write @playwright/test files unless asked.CodexOfficial OpenAI
  2. 02Cursor browser QA skillsA family of skills built on Cursor's built-in browser and accessibility tree — visual QA, responsive testing, dark-mode testing, and accessibility auditing.CursorCommunity spencerpauly
  3. 03network-request-auditingUses the browser's network log to flag failed requests, slowness, duplicates, and risky payloads after real user interactions.CursorCommunity spencerpauly
  4. 04agent-compatibilityScores how well a repo holds up under agent workflows — startup, validation, and docs reliability — by pairing a published CLI scan with three focused review agents.CursorOfficial Cursor

Deployment and CI

  1. 01gh-fix-ciInspects failing PR checks with gh, fetches the GitHub Actions logs, summarizes the failure, then proposes a fix plan and implements it only after explicit approval.CodexOfficial OpenAI
  2. 02deploy-on-awsDeploys applications to AWS with architecture recommendations, cost estimates, and infrastructure-as-code — one plugin in AWS's own Agent Plugins repository.Claude CodeCodexCursorVendor AWS Labs

Performance and monitoring

  1. 01adding-analyticsAdds PostHog event tracking, page views, feature flags, and session replay to a web app.CursorCommunity spencerpauly
  2. 02PostHog skills (official)PostHog's own official skill collection for analytics integration, feature flags, error tracking, and LLM analytics, auto-generated from PostHog's docs on every release.Claude CodeVendor PostHog

Curation

Why this isn’t a bigger list.

A skill only gets a profile here if it has a real source repository or official doc, a capability we can describe from that source rather than guess from its name, and enough material to say something a listing page can’t — when to use it, when not to, and what it depends on.

That’s deliberately a small fraction of what’s out there. Registries built for install-time discovery are the right tool for browsing everything; this is the tool for deciding whether a specific skill is right for a specific agent and task before you install it. See where to find agent skills for the full landscape, including the registries this library doesn’t try to replace.