Core Idea
The author's design philosophy, what sets this apart from peer projects, and quick notes on the key workflows. Skill names mentioned in the text auto-link to their detail pages.
This repository is Anthropic's official demonstration of the Agent Skills standard — a collection of self-contained SKILL.md folders that Claude loads dynamically to perform specialized tasks repeatably.
Install via Claude Code plugin marketplace or use directly in Claude.ai. Skills trigger automatically by context (e.g. mention of .pdf), or users can invoke by name like "Use the PDF skill".
Skill Relationships
Skill → Skill usage relationships, with cross-layer references. Same color blocks indicate the same skill form.
Skill Composition
Each skill is designed for a specific kind of caller — some only fire when you name them, others load automatically when the model judges them relevant. The split below summarises both modes at a glance.
These skills only fire when you call them by name. The agent will not load them on its own — they stay inert until invoked.
These skills are loaded automatically by the agent when it judges the context relevant. You usually don't need to call them yourself.
Skill List
Creates original algorithmic art via p5.js with seeded randomness and interactive parameters. Method: craft an algorithmic philosophy .md, deduce a conceptual seed, then implement p5.js viewer from template. Output: philosophy doc + interactive HTML/JS generative art, enabling unique reproducible compositions without copying existing artists.
Applies Anthropic's official brand colors and typography to artifacts needing the company look-and-feel. Provides color values, font rules, and accent cycling via style guidelines. Output is on-brand, consistent visual formatting with fallback fonts.
Creates original visual art in .png and .pdf using a two-step design philosophy method. First authors a visual movement manifesto, then expresses it on canvas with bundled fonts and minimal text. Delivers museum-quality, meticulously crafted static designs avoiding copyright infringement.
Helps build, modify, and migrate LLM-powered apps with the Claude API across Python, TypeScript, Java, Go, Ruby, C#, PHP, and cURL. Detects project language, reads language-specific docs, enforces official SDK usage, applies model defaults, and avoids API drift. Produces correct, up-to-date integration code and migration guidance.
Guides users through a structured 3-stage workflow (context gathering, refinement, reader testing) for co-authoring docs like specs and proposals. Uses clarifying questions, brainstorming, and iterative section drafting to transfer context and refine content. Produces well-structured documents that work for readers and catch blind spots early.
Skill for creating, reading, editing, and manipulating Word .docx/.dotx files via docx-js scripts, unzip/edit/zip workflows, pandoc, and LibreOffice validation. Produces polished, schema-valid Word documents with TOC, tables, tracked changes, or comments, visually verified by PDF rendering.
Guidance for distinctive, intentional frontend visual design when building or reshaping UI. Method: act as design lead, ground in subject, build token system, critique against AI defaults, iterate plan-build-critique. Value: helps produce non-templated, aesthetic, accessible interfaces with memorable signature elements.
A skill that helps write all kinds of internal communications using the company's preferred formats. It identifies the comms type, loads the matching guideline from examples/, and follows its formatting and tone rules. Produces properly structured status reports, newsletters, FAQs, and more that align with internal standards.
Guide for building high-quality MCP servers that let LLMs use external services via tools. Uses 4-phase workflow (research, implement, test, evaluate) with TypeScript/Python SDKs, best-practice refs, and eval scripts. Produces a working server and verification QA pairs.
A skill for any PDF task: extract text/tables, merge/split/rotate, watermark, encrypt, fill forms, OCR scans, and create PDFs via Python libs (pypdf, pdfplumber, reportlab) and CLI tools (qpdf, pdftk, poppler). It delivers processed files or extracted data efficiently.
Skill for creating, editing, and analyzing PowerPoint .pptx/.potx files via pptxgenjs scripts, XML unzip-edit-zip workflows, and helper scripts. Produces valid decks, extracts content, and applies design ideas, with schema validation for quality output.
A skill for creating new skills and iteratively improving them via drafting, testing with eval prompts, and benchmarking. It guides intent capture, SKILL.md authoring, parallel subagent runs, and quantitative/qualitative evaluation using bundled scripts. Output is a validated, optimized skill with better trigger accuracy and measured performance.
A toolkit for creating Slack-optimized animated GIFs using Python PIL. Provides Slack constraints, validation tools, easing and frame helpers, and animation concepts. Output is a polished, file-size-compliant GIF ready for Slack emoji or messages.
A template skill providing a minimal SKILL.md scaffold for creating new Agent Skills. It offers placeholder structure and guidance to fill in description and content, enabling quick bootstrapping of skill files with consistent format.
Toolkit for styling artifacts (slides, docs, HTML, reports) with 10 preset professional themes or custom on-the-fly themes. Method: preview PDF, select theme, apply colors/fonts from spec. Value: consistent, polished visual identity with minimal effort.
Suite of tools for building elaborate multi-component claude.ai HTML artifacts using React, Tailwind, and shadcn/ui. It scaffolds via init script, develops with modern frontend stack, and bundles to a single HTML file via bash scripts. Produces self-contained interactive artifacts shareable in chat, avoiding generic AI design.
Toolkit for testing local web apps via native Python Playwright scripts. Uses helper scripts for server management and a reconnaissance-then-action pattern. Enables verifying frontend functionality, debugging UI, capturing screenshots, and viewing browser logs.
Skill for creating, editing, and analyzing spreadsheet files (.xlsx, .csv, etc.) using Python libraries openpyxl, pandas, and markitdown. Key methods: write formulas via openpyxl, bulk data with pandas, recalculate with LibreOffice via recalc.py. Produces error-free, professionally formatted spreadsheets with validated formulas and clean structures.
Installation
Copy and run in your terminal to add anthropics/skills to your Agent environment.
npx skills add anthropics/skills