Project Rules MCP Server

An MCP server implementation that generates comprehensive project rules and best practices prompts for AI coding assistants like Cursor.

Features

• Generate high-quality analysis prompts for project code review
• Identify project purpose, scope, and boundaries
• Analyze technology stack and frameworks
• Document code style and formatting conventions
• Map architecture and structural patterns
• Identify components and API design patterns
• Document build and deployment processes
• Support both single-file and multi-file output formats
• Customizable analysis sections

Tool

generate_project_rules

Creates comprehensive prompts that guide AI tools to analyze your project and generate specific rules based on your codebase patterns.

Inputs:

• outputPath (string, optional): Output directory path (default: .pingancoder/rules)
• format (string, optional): Output format - "single" or "multiple" (default: "single")
• includeProjectPurpose (boolean, optional): Include project purpose analysis (default: true)
• includeTechStack (boolean, optional): Include technology stack analysis (default: true)
• includeCodeStyle (boolean, optional): Include code style analysis (default: true)
• includeArchitecture (boolean, optional): Include architecture analysis (default: true)
• includeComponents (boolean, optional): Include components analysis (default: true)
• includeBuildSteps (boolean, optional): Include build/deployment analysis (default: true)

Output:

• Creates detailed analysis prompts in markdown format
• Prompts include instructions for AI assistants to save results in .pingancoder/rules/ directory
• No files are created by the MCP tool itself - it only generates the analysis prompts

Usage

The Project Rules tool generates comprehensive analysis prompts that guide AI assistants to:

• Generate project-specific analysis prompts
• Help AI assistants understand your codebase patterns
• Create consistent coding guidelines across your project
• Maintain architectural consistency
• Provide detailed instructions for code analysis

Workflow

1. Generate Analysis Prompts: Use the MCP tool to create comprehensive analysis prompts
2. Copy Prompts to Cursor: Copy the generated prompts and paste them into Cursor
3. Run Analysis: Cursor will scan your codebase and create specific rules based on the prompts
4. Save Rules: Cursor will automatically save the generated rules in the .pingancoder/rules/ directory

How to Use

1. Generate prompts using the MCP tool
2. Copy the generated prompts from the tool output
3. Paste the prompts into Cursor and run the analysis
4. Cursor will automatically save the analysis results in your project's .pingancoder/rules/ directory

The prompts include specific instructions for where to save the generated rules, so you don't need to manually specify the output location.

Configuration

Usage with Claude Desktop

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "project-rules": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-project-rules"
      ]
    }
  }
}

Usage with VS Code

For manual installation, add the following JSON block to your User Settings (JSON) file in VS Code:

{
  "mcp": {
    "servers": {
      "project-rules": {
        "command": "npx",
        "args": [
          "-y",
          "@modelcontextprotocol/server-project-rules"
        ]
      }
    }
  }
}

Local Development

For development, you can run the server locally:

npm run build
node dist/index.js

Output Structure

The tool generates analysis prompts in markdown format that instruct AI assistants to save their analysis results in the .pingancoder/rules/ directory:

Single Format (format: "single")

• Generates one comprehensive prompt covering all analysis areas
• AI will save results as .pingancoder/rules/project-rules.md

Multiple Format (format: "multiple")

• Generates separate specialized prompts for each analysis area
• AI will save results as separate files:
  • .pingancoder/rules/project-rules.md - Overall project analysis
  • .pingancoder/rules/tech-stack.md - Technology stack analysis
  • .pingancoder/rules/code-style.md - Code style and standards
  • .pingancoder/rules/architecture.md - Architecture and structure
  • .pingancoder/rules/components.md - Components and APIs
  • .pingancoder/rules/build-deploy.md - Build and deployment

Example Usage

Generate Single Comprehensive Prompt

echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"generate_project_rules","arguments":{"format":"single"}}}' | node dist/index.js

Generate Multiple Specialized Prompts

echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"generate_project_rules","arguments":{"format":"multiple"}}}' | node dist/index.js

Generate with Custom Sections

echo '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"generate_project_rules","arguments":{"includeBuildSteps":false,"includeArchitecture":false}}}' | node dist/index.js

Usage with Cursor

1. Run one of the above commands to generate analysis prompts
2. Copy the generated prompt text from the output
3. Paste the prompt into Cursor
4. Cursor will analyze your project and save the results in .pingancoder/rules/

Development

Building

npm run build

Testing

npm test

License

This MCP server is licensed under the MIT License. This means you are free to use, modify, and distribute the software, subject to the terms and conditions of the MIT License.