Cursor 安装指南
常与开发类 MCP Server 搭配使用的 AI 代码编辑器。
MCP Servers
It's been a constant struggle trying to understand unfamiliar and complex codebases - managing cognitive overload and trying to imagine how everything fits together.
The QuantConnect MCP Server is a bridge for AIs (such as Claude and OpenAI o3 Pro) to interact with our cloud platform.
FreeCAD 3D CAD automation via FastMCP — headless document/export plus FluidX3D/OpenFOAM CFD extensions. For Cursor / Claude Desktop. Webapp.
A retrieval layer for AI agents over peer-reviewed papers, books, patents, standards, Wikipedia, Reddit, Telegram, Discord, and YouTube.
A datetime formatting service implemented as an MCP server for the Claude Desktop Application. Supports generation of datetime strings in various formats.
Stream Deck MCP lets agents build and reconfigure real Elgato Stream Deck profiles.
MCP server for AI-powered UI component discovery on ui-layouts.com — search, docs, metadata, and TSX source in one pipeline.
Define TypeScript capabilities that send emails, manage calendars, and automate work. Expose them to the Routecraft agent, Claude, ChatGPT, Cursor, or any AI agent via MCP.
MCP-powered NetEase Cloud Music player: play music from NetEase Cloud Music via neteasecli and mpv, with a local synced-lyrics web player.
Local MCP server for COMSOL modeling workflows, validation, and RAG-assisted multiphysics automation.
The watchman for your local AI agents' MCP layer.
MCP server for Taiwan's largest job platform — search jobs, browse companies, and apply, all from Claude.
Agent Skills
This project is under active development and evolving rapidly. Contributions, feedback, and discussions are warmly welcome.
Agent skill that audits vibe-coded apps for common security vulnerabilities introduced by AI coding assistants
A high-performance Markdown parser and renderer for Angular, React, Svelte, Vue, HTML and ANSI.
A collection of Codex/agent skills for planning, documentation access, frontend development, and browser automation.
简单说，就是跟 Claude 说一句话，它会帮我从每天的新论文里筛一轮，挑出值得看的，再把重点论文读完、写成 Obsidian 笔记。日常不用记一堆命令，基本就是：