도구

CLI and TypeScript runtime for discovering and calling MCP servers from the terminal and scripts: list tool signatures and call tools directly, or generate a standalone CLI or typed client from a server. Auto-imports existing MCP configs from Claude Code, Codex and others.

Reads any website's design system in a headless browser and exports DTCG design tokens, Tailwind config, Figma variables and a brand guide in one command, scoring contrast and cross-page consistency. Also an MCP server or Claude Code plugin. Needs Node.js and Chromium or Chrome.

Runs Google Lighthouse across an entire site (sampling similar pages smartly) and summarizes SEO, performance and accessibility scores page by page in a local web UI; unlighthouse-ci writes reports directly for agents to read. Great for site-wide SEO checkups. Needs Node.js 22.18+.

Design token build tool: define colors, font sizes, spacing and other design specs in one JSON source, then style-dictionary build generates CSS, SCSS, iOS, Android and other platform code, keeping the design system and implementations in sync. Needs Node.js and a config.json.

Draws bar charts, histograms, line charts, scatter plots and box plots in the terminal from CSV/TSV on stdin, so you can glance at data distributions at the end of a shell pipeline. The command is uplot.

Teaches agents to write video prompts for ByteDance's Seedance 2.0: input limits, @ asset references and camera language, with templates for product ads, short dramas, dance, scenic clips, e-commerce and explainers, in English and Chinese. Prompts only; generate videos with your own account.

Teaches agents to code against the latest Gemini API and Python/TypeScript SDKs: multi-turn chat, function calling, structured output, and image and speech generation, while avoiding deprecated older models. Maintained by Google's Gemini team.

Gemini API development skills: gemini-api-dev (core API and SDKs), gemini-live-api-dev (real-time bidirectional audio/video streaming) and gemini-omni-flash-api (Omni video generation and editing), filling in APIs released after model training cutoffs. Maintained by Google's Gemini team.

Visually compares two PDFs page by page: reports differences via exit code, writes a PDF with differences highlighted, or opens a GUI for side-by-side review. Useful for checking contract or generated-document revisions. No longer actively developed.

Multithreaded lossless PNG optimizer that shrinks screenshots, slices and icons in bulk without quality loss and strips unneeded metadata, handy for design handoff or compressing images before a site launch. Install with Homebrew on macOS.

High-fidelity SVG rendering library and CLI that exports SVG icons and illustrations to PNG at chosen sizes, good for batch asset export. No system dependencies, with pixel-identical output across platforms. No SVG animation support; install with cargo or grab a macOS build from Releases.

Fast async link checker written in Rust that scans websites and Markdown, HTML and other files for broken links and email addresses, with JSON output, config files and an official GitHub Action, often wired into docs CI.

Hugging Face's official hf CLI: search, download and upload models and datasets, manage repos and storage buckets, and run jobs on Hugging Face's cloud. Output adapts automatically for agents, and hf skills add installs companion skills. Uploads and similar actions need a login.

Extracts any website's logo, colors, fonts, spacing, shadows and motion in one command, exporting W3C design tokens, a Tailwind theme, DESIGN.md or a brand guide PDF, plus WCAG contrast checks. Good for competitor and brand audits. Needs Node.js 18+; run dembrandt install-browser first.

Design engineer Jakub Krehel's UI polish skill covering animation, typography, icons, hover states, optical alignment, concentric radii, shadows and hit areas, with quick and full review modes, so agents get the details right when building or reviewing UI. Works on web code; needs Node.js.

NVIDIA's official plugin that finds and installs the right skills from NVIDIA's catalog for your task, spanning CUDA acceleration, NIM inference, NeMo training, RAG blueprints, Omniverse simulation and robotics. Listed in both the Claude and Codex official marketplaces.

NVIDIA's official, verified catalog of hundreds of skills covering CUDA-X libraries, NeMo training, NIM and Dynamo inference, RAG blueprints, Jetson, Omniverse and robotics simulation. The install adds only nvidia-skill-finder, which recommends other skills and installs them once you confirm.

Fast Rust toolkit for tabular data with nearly 80 subcommands for querying, stats, cleaning, validation, joins and format conversion on CSV and Excel. Also offers an MCP server and a Claude Cowork plugin.

Browserbase's official skills: drive a local or cloud browser with the browse CLI for web scraping and search, turning site traffic into an OpenAPI spec, adversarial UI testing and competitor research. Cloud features need a Browserbase account.

Official Scenario skills: generate production-ready images, video, voiceover and music, textures and 3D assets through the Scenario MCP, for product shots, ads, storyboards and game art, plus expert skills for Blender, Maya and ZBrush. Needs a Scenario account (free tier available).

AWS's generally available agent toolkit and plugin marketplace: plugins such as aws-core, aws-agents and aws-data-analytics bundle the AWS MCP Server with evaluated skills for IaC, serverless, containers and billing. Calling AWS APIs requires configured credentials.

PostgreSQL MCP server with database health checks, index tuning, EXPLAIN plan analysis and hypothetical index simulation, plus safe read-only SQL execution in restricted mode.

Teaches agents to build, debug and review stateful AI agents with the Cloudflare Agents SDK: state sync, scheduled tasks, RPC, MCP servers, email and streaming chat, checking the latest official docs first. Maintained by Cloudflare.

Claude Code plugin marketplace for designers: 33 plugins and 273 skills for user research, design systems, UX strategy, interaction, prototype testing, visual critique and accessibility. Install designer-toolkit and run /designer-toolkit:start-here for a suggested flow; no coding needed.