MCP Pattern Language

🚀 My Rules

An MCP server that stores all rules, prompts, etc., enabling agents to call rules on - demand.

🚀 Quick Start

✨ Features

This MCP server is conceptually similar to Cursor's rules. It allows coding agents to look up rules within the context of their current tasks, eliminating the hassle of carrying small notes across projects and IDEs and reducing the need for manual rule invocation.

📦 Installation

Claude Code

claude mcp add --transport http rules https://mcp.cianfrani.dev/mcp

Cursor

Claude Desktop

{
    "rules": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.cianfrani.dev/mcp"
      ]
    }
}

💻 Usage Examples

Basic Usage

The main function of this server is to allow coding agents to look up rules on - demand. For example, when writing a test for @calc.py:

> write a test for @calc.py
  ⎿  Read calc.py (26 lines)

⏺ rules:get_rules_for_writing_tests (MCP)(language: "python")
  ⎿  - Before testing a protected method, ask "Can this logic be adequately tested through the public interface?". If yes, don't test the protected method directly.
     If no, continue testing the protected method.
     - Remember to write tests using the Arrange, Act, Assert pattern.
     … +12 lines (ctrl+r to expand)

⏺ Write(test_calc.py)

📚 Documentation

What's It Do?

It enables coding agents to look up rules as needed within the context of their current tasks.

Why?

• Carrying small notes across projects and IDEs is extremely annoying.
• Manual rule invocation is not desired.

How's It Work?

Well - defined tool descriptions allow the agent to fetch relevant rules on - demand.

What Doesn't work?

• Patterns must be associated with specific actions, such as "writing tests", "authoring UI", "reviewing a PR".
• The agent decides whether to call the tool, and the result is not always predictable.

Tools

Prompts

In Claude Code, invoke prompts as slash commands.

📚 Resources

• Deploying this shit on netlify

🔧 Technical Details

This section could potentially cover more in - depth implementation details, such as how the server communicates with the agents, the data storage mechanism for rules, etc. However, the original README does not provide sufficient content for a detailed technical description.

🚧 Dev

Start the server

npm run dev

Open MCP inspector

npm run inspect