A community-driven registry for Claude, Cursor, Windsurf, Cline & more. Not affiliated with Anthropic.
Are you the author? Sign in to claim
Bootstrap of Cursor rules (.mdc) and mirror documentation (.md) for AI agent-guided projects.
CLI tool to bootstrap AI agent rules for multiple IDEs and frameworks. Generates optimized rules for Cursor, VS Code, Claude, and 6+ other AI coding assistants.
Choose from 9 supported IDEs including Cursor, VS Code/GitHub Copilot, Windsurf, Continue, Claude, and more.
Agent Rules Kit is a CLI tool that facilitates the installation and configuration of rules for multiple AI-powered IDEs and coding assistants. The rules help AI agents better understand the structure, patterns, and best practices of different technology stacks, generating optimized configurations for each supported IDE.
Listen to a brief introduction about Agent Rules Kit:
https://github.com/user-attachments/assets/7d65c696-245d-421d-9ddc-90331a92c9b2
https://github.com/user-attachments/assets/8e91d651-c15f-4892-a250-684ab60d8594
# Install and run interactively
npx agent-rules-kit
# The CLI will guide you through:
# 1. 📁 Project path selection
# 2. 🎯 IDE selection (Cursor, VS Code, Claude, etc.)
# 3. 📚 Stack and architecture choice (optional)
# 4. 🌐 Global best practices (recommended)
# 5. 🔧 MCP tools selection (multiple tools supported)
# 6. ⚡ Automatic rule generation
For automated or scripted installations, use command line arguments to skip interactive prompts:
# Basic installation with auto-detection
npx agent-rules-kit --stack=laravel --version=12 --global --auto-install
# Full configuration example
npx agent-rules-kit --stack=nextjs --version=14 --architecture=app --global --mcp-tools=pampa,github --ide=cursor
# For different IDEs
npx agent-rules-kit --stack=react --version=18 --ide=vscode # VS Code
npx agent-rules-kit --stack=laravel --version=12 --ide=claude # Claude
npx agent-rules-kit --stack=angular --version=17 --ide=windsurf # Windsurf
📖 For complete CLI options and examples, see AGENTS.md
Your rules will be generated in the appropriate location for your chosen IDE and automatically detected!
If you are using an AI agent, check AGENTS.md for comprehensive instructions on automated installation. Share this file with your agent so it installs the rules automatically with the correct options.
We've completely redesigned the internal architecture to provide a more maintainable and extensible system:
Since v2.0: Agent Rules Kit includes comprehensive support for Model Context Protocol (MCP) tools! This powerful feature allows AI agents to work more effectively with external tools and services.
Model Context Protocol (MCP) tools are standardized interfaces that allow AI assistants to securely connect to external data sources and services. These tools enable AI agents to:
Choose from 5 popular MCP tools during installation:
| Tool | Name | Description | Best For |
|---|---|---|---|
| 🔍 PAMPA | Semantic Code Search | AI-powered code search with project memory | Understanding large codebases, finding relevant code patterns |
| 🐙 GitHub | Repository Management | Secure GitHub API integration for file operations | Repository management, collaborative development |
| 🧠 Memory | Persistent Knowledge | Long-term memory storage across sessions | Maintaining context, learning from previous interactions |
| 📁 Filesystem | File Operations | Secure file system access with controls | File management, project organization |
| 🔀 Git | Version Control | Repository operations and history analysis | Version control, commit management, branch operations |
The new multi-select interface allows you to install rules for multiple MCP tools simultaneously:
? Select MCP tools to install rules for: (Use space to select, enter to confirm)
❯◉ PAMPA - Semantic Code Search - AI-powered semantic code search and project memory system
◉ GitHub - Repository Management - Secure access to GitHub repositories for file operations
◯ Memory - Persistent Knowledge - Persistent knowledge storage and retrieval across sessions
◉ Filesystem - File Operations - Secure file operations with configurable access controls
◯ Git - Version Control - Repository operations, commit history analysis, and version control
Each selected MCP tool generates organized rules in your project:
.cursor/rules/rules-kit/mcp-tools/
├── pampa/
│ ├── pampa-mcp-usage.mdc # Core usage patterns
│ └── pampa-best-practices.mdc # Best practices and tips
├── github/
│ ├── github-mcp-usage.mdc # GitHub API integration rules
│ └── github-workflow.mdc # Workflow patterns
└── memory/
├── memory-mcp-usage.mdc # Memory management rules
└── memory-patterns.mdc # Common memory patterns
MCP Tools rules help AI agents:
MCP tools rules work alongside your stack-specific rules:
| Stack | Status | Features |
|---|---|---|
| Since v2.0: Multi-select MCP tools (PAMPA, GitHub, Memory, Filesystem, Git), usage patterns, security guidelines, workflows | ||
| Complete SDK coverage (Python, TypeScript, Java, Kotlin, C#, Swift), server/client patterns, architecture concepts | ||
| Multiple architectures, version detection, casting rules | ||
| Testing guidelines, architectural patterns, version-specific docs | ||
| Semantic code search, AI agent integration, MCP support, multi-language indexing | ||
| Architecture options, state management, testing guidelines, version-specific guidance (v18 concurrent features) | ||
| Base rules (architecture, best practices, styling, naming, navigation, platform code, state, testing, version info) | ||
| App & Pages router support, version detection, testing docs, version-specific guidance (v13 API routes, v14 Server Components) | ||
| Component organization, runes (v5), lifecycle, state management | ||
| Multiple architectures, project structure, error handling patterns | ||
| Routing system, layouts, SSR/CSR strategies, form actions | ||
| Content collections, static/dynamic content, integration guides | ||
| Signals support, base project structure | ||
| Standard and Microservices architectures, NestJS 9.x and 10.x features | ||
| Base project structure, best practices, naming, testing | ||
| Testing guidelines, architecture concepts |
Enhanced Extensibility: Detailed documentation has been added on how to extend Agent Rules Kit with new services, making it easier to contribute new stacks and features.
For guidelines on contributing to this project, please see CONTRIBUTING.md.
Please read our Code of Conduct before participating in our project.
This project is licensed under the ISC License. See the LICENSE file for details.
Choose from a wide range of technology stacks:
| Stack | SDKs/Versions | Architectures |
|---|---|---|
| Laravel | 8, 9, 10, 11, 12 | Standard, DDD, Hexagonal |
| Next.js | 12, 13, 14 | App Router, Pages Router |
| Angular | 14, 15, 16, 17 | Standard, Standalone, Micro-frontends |
| React | 17, 18 | Standard, Hooks, Concurrent |
| Vue.js | 2, 3 | Options API, Composition API, Nuxt |
| Astro | 2, 3, 4, 5 | Static, SSR, Hybrid |
| React Native | 0.70, 0.71, 0.72, 0.73, 0.74 | Standard, Expo |
| NestJS | 8, 9, 10 | Standard, Microservices, GraphQL |
| Go | 1.20, 1.21, 1.22 | Standard, DDD, Hexagonal |
| Node.js | 18, 20 | Standard |
| 🆕 MCP (Model Context Protocol) | Python, TypeScript, Java, Kotlin, C#, Swift | Server, Client, Toolkit |
| 🆕 Pampa | Latest | Standard |
MCP is an open standard that enables AI applications to securely connect to data sources. Our rules kit includes comprehensive implementation guides for all major SDKs:
Pampa is our own AI-powered semantic code search tool designed specifically for AI agents:
Perfect for AI agents to quickly understand and navigate any codebase!
Agent Rules Kit can install existing rules into several IDEs and agents using the new install command. Pass the desired IDE with --target (or the alias --ide) to skip the interactive selector:
npx agent-rules-kit install --target=cursor # or --ide=cursor
| Target | Path | Extension | Front matter |
|---|---|---|---|
| Cursor | .cursor/rules/ | .mdc | yes |
| VS Code / GitHub Copilot | .github/copilot-instructions.md | .md | no |
| Windsurf | .windsurf/rules/ | .md | no |
| Windsurf (legacy) | .windsurfrules | .md | no |
| Continue | .continue/rules/ | .md | yes |
| Zed | ./.rules | .rules | no |
| Claude Code | CLAUDE.md | .md | no |
| Gemini Code | GEMINI.md | .md | no |
| OpenAI Codex | AGENTS.md | .md | no |
| Cline | .clinerules | .md | no |
The command detects .mdc files in your project (default ./rules) and writes them in the format required by the chosen target.
A scaffold demonstrating how to use a turbo, mono repo, trpc, better auth, react, postgres and cursor ai rules.
A practical approach to managing multiple AI agents in Cursor through strict file-tree partitioning and domain boundarie
📄 Configuration files that enhance Cursor AI editor experience with custom rules and behaviors