Skills by Anthropic

Guides documentation from context gathering through section-by-section drafting and reader testing.

Create, read, edit, validate, and visually inspect .docx documents.

Read, create, extract, merge, split, rotate, watermark, OCR, and protect PDF files.

Create, inspect, edit, and validate PPTX decks and POTX templates.

Create, edit, inspect, and validate spreadsheet files while preserving formulas, formatting, and workbook conventions.

Create a seeded p5.js artwork, its algorithmic manifesto, and an interactive parameter viewer.

Create and refine agent skills through structured drafting, testing, comparison, and user review.

Creates an original visual philosophy and renders it as a refined PNG or PDF artwork.

Build and evaluate TypeScript or Python MCP servers that expose external services as well-designed tools.

Create and validate Slack-ready animated GIFs with PIL-based drawing, motion helpers, and size optimization.

Plan and build subject-specific interfaces with deliberate color, typography, layout, motion, and copy.

Build Claude API integrations with documented, language-specific SDK patterns and current model defaults.

Test local web apps with Playwright scripts, managed servers, DOM inspection, screenshots, and browser logs.

Build React artifacts with a configured UI stack and bundle them into one self-contained HTML file.

Apply a preset or custom color-and-font theme consistently across a slide deck or other artifact.

Apply Anthropic’s official colors, heading fonts, body fonts, and shape accents to visual artifacts.

Draft internal updates using the appropriate format, tone, and content-gathering guidelines.

>

>