Agent Skills Directory for Claude, Codex and Cursor

Recommends matching courses, tutorials, and use cases from Claude Academy when you ask how to use Claude or its features.

Creates original generative art with p5.js: an algorithmic philosophy document plus a self-contained interactive HTML viewer with seeded randomness and sliders.

Applies Anthropic's official brand colors and typography to documents and slides so they match the company's visual style.

Create original visual art, posters and designs as .png or .pdf files, guided by a custom design philosophy.

Helps build and maintain applications that call the Claude API, with up-to-date model IDs, pricing, parameters, streaming, tool use, and migration guidance.

Adds 2-3 short follow-up questions after substantive answers, helping users check facts, assumptions, and missing context before acting.

Guides you through a three-stage workflow to co-author docs, specs, proposals and decision documents with iterative refinement.

Lets the assistant create, read, edit, and convert Word documents (.docx), including tracked changes, comments, and templates.

Guides the assistant to design distinctive, non-templated visual interfaces with intentional typography, color, layout, and copy.

Helps write internal company communications like 3P updates, newsletters, FAQs, status reports, and incident reports using your company's preferred formats.

Guides the assistant through building high-quality MCP servers in Python or TypeScript, from planning and coding to testing and evaluations.

Lets the assistant work with PDF files: extract text and tables, merge, split, rotate, watermark, create, fill forms, encrypt, and OCR scanned documents.

Create, edit, read and check PowerPoint (.pptx) presentations, including templates, charts, images and speaker notes.

Guides the assistant through creating, testing, evaluating, and improving new skills, including optimizing their descriptions for better triggering.

Helps create animated GIFs sized and optimized for Slack, including emoji GIFs, using Python drawing and animation tools.

Applies one of ten pre-set color and font themes, or a custom-generated theme, to slides, documents, reports and HTML pages.

Builds elaborate multi-component claude.ai artifacts with React, Tailwind and shadcn/ui, bundled into a single HTML file.

Lets the assistant test local web apps by writing and running Playwright scripts to click through pages, take screenshots, and read browser logs.

Lets the assistant create, edit, fix, and analyze Excel and CSV spreadsheet files, including formulas, formatting, charts, and data cleaning.

Guides the assistant through a structured four-phase self-debugging workflow when an agent run fails, loops, or burns tokens without progress.

Analyzes a code repository and sorts skills, commands, rules, and hooks into DAILY vs LIBRARY buckets to build a trimmed, evidence-backed install plan.

Guides AI-assisted engineering with eval-first workflows, task decomposition, cost-aware model routing, and focused code review practices.

Guides the assistant to design and review REST APIs using best practices for naming, status codes, pagination, errors, versioning, and rate limiting.

Guides the assistant in designing consistent, production-ready REST APIs following naming, status code, pagination, error, and versioning best practices.