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
Outcome-Based Persistent Memory MCP Server
The open-source trust layer for bringing Claude Code (and other autonomous AI agents) to work — with your security team's approval.
A self-modifying, self-verifying AI agent with persistent state. Reads its own source code. Plans changes. Tests them in a sandbox before applying.
Open security engineering for code written by humans and AI agents. Discover. Repair. Verify.
Local-first knowledge compiler for humans and agents. Inspired by Karpathy's vision. Plain markdown, CLI + MCP server.
Open-source, local-first memory and context for AI agents. One memory store, every agent.
Hosted uptime monitoring and public status pages for teams. Multi-region checks, automatic incidents, on-call alerts (Slack/Telegram/webhook/+), subscribers, REST API, Terraform provider, MCP server.
An MCP server that connects to a Swagger specification and helps an AI to build all the required models to generate a MCP server for that service.
RimBridgeServer runs an MCP server inside RimWorld so AI agents and external tools can remotely control and observe a running game.
Use Devnors Data from Python with one API Key.
ChatCrystal is a local-first AI PKM app for developers who solve real problems with Claude Code, Cursor, Codex CLI, Trae, and GitHub Copilot.
⚠️ This implementation stores OAuth credentials in ./mnt/mcp-creds/ which are currently committed to the repository. Proper OAuth flow for MCP servers is coming soon to the Agencii platform.
Agent Skills
华创金工形态学Skill
Track your AI coding tool token usage and sync to vibecafe.ai.
— 此 skill 定義了「完成任務後必須詢問下一步」的強制模式，不可跳過、不可例外。
一个真正适合职场人的 PPT Skill。把文档丢给你的 AI Agent，每一页都自带编辑控制台的 PPT Skill——不满意的地方直接在浏览器里改，改完还能一键导出成真实的、可编辑的 PPTX。
A prose-native implementation of Vincent Driessen's git-flow branching model.
Claude Code version of skill-autoimprove — auto-improve any CLAUDE.md using Karpathy autoresearch loop
“他们都叫我小唐人！他们都叫我臭唐氏！说谁是唐氏呢！！！骂谁唐氏！” “关注永雏塔菲喵，关注永雏塔菲谢谢喵！”
🎯 Goal-driven execution helper for long-running tasks and explicit success criteria
"你们搞大模型的就是码奸，你们已经害死前端兄弟了，还要害死后端兄弟……" ——某位对未来充满警惕的工程师
Claude Code skill to extract React components from any production website
Claude Code skill for multi-reviewer peer review of academic papers. Adapted from poldrack/ai-peer-review — uses parallel Claude subagents instead of multiple proprietary LLMs.
这是一套专业的 Claude Skills，帮助开发者**源代码，而不只是"看懂"。基于认知科学研究，融合精细询问法、自我解释测试和应用迁移验证，确保深度学习而非产生流畅幻觉。