agent-rules-mcp

agent-rules-mcp

MCP server that enables your agents to use coding rules from any or your GitHub repository. Instead of workspace rules files, you can now prompt agents to access the your coding rules from any repository.

Category
访问服务器

README

Agent Rules MCP Server

MCP Server NPM Version MIT License

This MCP server eliminates the need for local rule files in your workspace. Instead of copying coding standards into each project, you can now prompt AI agents to fetch specific coding rules or all your rules from any rules folder on a public repository or your own.

Features

  • GitHub Integration: Fetches rules from any GitHub repository
  • Simple Setup: Configure with environment variables, no local files needed
  • Configurable: Support for custom repositories, branches, and paths
  • Community Rules: Works with existing collections like awesome-cursorrules and awesome-copilot, etc.
  • Compound Extensions: Supports .chatmode.md, .prompt.md, .instructions.md files
  • Flexible Format: Supports any markdown files (.md/.mdc) with or without metadata

MCP Client Configuration (default)

Add this configuration to your MCP client (VS Code, Kiro, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "agent-rules": {
      "command": "npx",
      "args": ["-y","agent-rules-mcp@latest"],
      "env": {
        "GITHUB_OWNER": "4regab",
        "GITHUB_REPO": "agent-rules-mcp",
        "GITHUB_PATH": "rules",
        "GITHUB_BRANCH": "master"
      },
      "disabled": false
    }
  }
}

Example Use of Community Rules Collections

GitHub Awesome Copilot Collection

Get instant access to community-maintained coding rules:

{
  "mcpServers": {
    "agent-rules": {
      "command": "npx",
      "args": ["-y","agent-rules-mcp@latest"],
      "env": {
        "GITHUB_OWNER": "github",
        "GITHUB_REPO": "awesome-copilot",         
        "GITHUB_PATH": "instructions",
        "GITHUB_BRANCH": "main"
      },
      "disabled": false
    }
  }
}

Awesome Cursor Rules Collection

Alternative collection for cursor-specific rules:

{
  "mcpServers": {
    "agent-rules": {
      "command": "npx",
      "args": ["-y","agent-rules-mcp@latest"],
      "env": {
         "GITHUB_OWNER": "PatrickJS",
         "GITHUB_REPO": "awesome-cursorrules",
         "GITHUB_PATH": "rules-new",
         "GITHUB_BRANCH": "main"
         }
       }
     }
   }

Available Tools

  • get_rules: Retrieves rule content for one or multiple domains from the GitHub repository rules folder.
  • list_rules: Lists all available rule domains with descriptions.

Using Your Own Rules Folder Repository (Recommended)

To use your own GitHub repository instead of the default:

{
  "mcpServers": {
    "agentrules": {
      "command": "npx",
      "args": ["-y","agent-rules-mcp@latest"],
      "env": {
        "GITHUB_OWNER": "your-username",
        "GITHUB_REPO": "your-rules-repo",
        "GITHUB_PATH": "your-rules-folder",
        "GITHUB_BRANCH": "main"
      },
      "disabled": false
    }
  }
}

Example repository structure:

my-coding-rules/
├── rules/                       # Traditional single directory
│   ├── python-style.md          # Standard markdown with metadata
│   ├── react-patterns.mdc       # MDC format supported
│   └── security-checklist.md    # With YAML frontmatter
├── README.md
└── .gitignore

How This Helps

On-Demand Rule Access for AI Agents

Before (Traditional Approach):

my-project/
├──rules                  ← Local rule files needed
│   ├── react-rules.md
│   ├── security-rules.md
│   └── typescript-rules.md
├── src/
└── package.json

After (agent-rules MCP Approach):

my-project/
├── src/
└── package.json          ← Clean workspace, no local rules needed

# In Coding Agent:
"Apply React best practices to this component"
→ Agent automatically fetches React rules from your rules folder

Flexible Support & File Format Compatibility

The server works with various file formats and naming conventions:

Supported Extensions:

  • .md - Standard markdown files
  • .mdc - MDC (Markdown Components) files
  • .chatmode.md - AI assistant mode definitions
  • .prompt.md - Prompt templates
  • .instructions.md - Coding instruction files

Automatic Metadata Extraction: If no explicit metadata is provided, the server will:

  • Extract the first heading as a title
  • Use the first paragraph as a description
  • Generate a fallback description based on the filename
  • Parse YAML frontmatter when available

Domain Name Handling:

  • accessibility.chatmode.md → domain: accessibility
  • react-best-practices.instructions.md → domain: react-best-practices
  • 4.1-Beast.chatmode.md → domain: 4.1-Beast (supports dots and special chars)

This means you can use any existing markdown documentation as rules without modification.

Contributing

We welcome contributions to the default rule repository!

  • Clear Domain Names: Use descriptive, kebab-case filenames
  • Complete Metadata: Include description and last updated date
  • Quality Content: Provide actionable, well-organized rules with examples
  • Test Locally: Verify your rules work with the MCP server
  • Follow Format: Use standard markdown structure

Recommended Structure (for optimal metadata extraction):

# Title of the coding rules

- Last Updated: YYYY-MM-DD
- Description: Brief description of the rules (used in list_rules() responses)
- Version: X.X (optional, for tracking major changes)

## Content 

License

MIT License - see LICENSE file for details.

Support

  • Issues: Report bugs and feature requests on GitHub Issues
  • Documentation: Check this README and inline code documentation

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选