codesona-mcp-server
Provides team coding standards to AI assistants via the MCP protocol, with tools to fetch rules and suggest new ones.
README
Codesona MCP Server
A standard MCP server that provides team coding standards via stdio transport. This server connects to the Codesona API and serves your team's coding rules to AI assistants in IDEs like Cursor, Windsurf, and other MCP-compatible tools.
Features
- 🔄 Automatic rule fetching from centralized Codesona API
- 📦 Smart caching (5-minute cache)
- ⚡ Fast & lightweight Node.js implementation
- 🔐 Secure API key authentication
- 📊 Full MCP protocol compliance
Installation
Prerequisites
- Node.js 18.0.0 or higher
- A Codesona workspace API key (get it from Codesona dashboard)
Method 1: Using npx (Recommended - No Installation Required)
This is the easiest method. No global installation needed.
-
Get your API key from Codesona dashboard
-
Configure your IDE (see IDE-specific instructions below)
Method 2: Global Installation
Install the package globally for system-wide access:
npm install -g @codesona/mcp-server
After installation, configure your IDE with the codesona-mcp-server command.
Method 3: From Source
- Clone the repository
git clone https://dev.azure.com/Azteron/Codesona/_git/CodesonaMCPServer
cd CodesonaMCPServer
- Install dependencies
npm install
- Make executable (Unix/Mac only)
chmod +x index.js
- Configure your IDE using the full path to
index.js
Configuration
Environment Variables
CODESONA_API_KEY(required): Your workspace API key from Codesona dashboardCODESONA_API_URL(optional): Codesona API URL (default:https:/codesona.ai)NODE_ENV(optional): Environment mode (development/production)
IDE Setup Instructions
Cursor
-
Open Cursor Settings
- Mac:
Cmd + ,orCode > Settings - Windows/Linux:
Ctrl + ,orFile > Preferences > Settings
- Mac:
-
Find MCP Settings
- Search for "MCP" in settings search bar
- Or navigate to: Extensions > MCP
-
Edit MCP Configuration
- Click "Edit in settings.json"
- Add the following configuration:
Using npx (Recommended):
{
"mcpServers": {
"codesona": {
"command": "npx",
"args": ["-y", "@codesona/mcp-server"],
"env": {
"CODESONA_API_KEY": "your-workspace-api-key-here"
}
}
}
}
Using global installation:
{
"mcpServers": {
"codesona": {
"command": "codesona-mcp-server",
"env": {
"CODESONA_API_KEY": "your-workspace-api-key-here"
}
}
}
}
-
Replace the API key
- Change
your-workspace-api-key-hereto your actual API key from Codesona dashboard
- Change
-
Save and restart Cursor
- Save the file (
Cmd+S/Ctrl+S) - Completely restart Cursor (not just reload window)
- Save the file (
Windsurf
-
Open Windsurf Settings
- Navigate to Settings
-
Find MCP Configuration
- Look for MCP Server settings
- Click "Configure MCP Servers"
-
Add Codesona Server
Using npx (Recommended):
{
"mcpServers": {
"codesona": {
"command": "npx",
"args": ["-y", "@codesona/mcp-server"],
"env": {
"CODESONA_API_KEY": "your-workspace-api-key-here"
}
}
}
}
Using global installation:
{
"mcpServers": {
"codesona": {
"command": "codesona-mcp-server",
"env": {
"CODESONA_API_KEY": "your-workspace-api-key-here"
}
}
}
}
-
Replace the API key and save
-
Restart Windsurf
Claude Desktop
-
Open Claude Desktop Configuration
- Mac:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
- Mac:
-
Add MCP Server Configuration
Using npx (Recommended):
{
"mcpServers": {
"codesona": {
"command": "npx",
"args": ["-y", "@codesona/mcp-server"],
"env": {
"CODESONA_API_KEY": "your-workspace-api-key-here"
}
}
}
}
Using global installation:
{
"mcpServers": {
"codesona": {
"command": "codesona-mcp-server",
"env": {
"CODESONA_API_KEY": "your-workspace-api-key-here"
}
}
}
}
-
Replace the API key and save
-
Restart Claude Desktop
Available MCP Tools
get_coding_rules
Retrieves mandatory team coding standards. Accepts optional codeContext parameter:
language: Programming language (e.g., "csharp", "javascript")framework: Framework name (e.g., "aspnet", "react")platform: Platform type (e.g., "backend", "frontend")
Rules are cached for 5 minutes.
suggest_new_rule
Suggests a new coding rule to team standards. Required parameters: title, description, rationale. Optional: exampleCode, badExampleCode, goodExampleCode, codeContext.
Available MCP Resources
- Team Coding Standards (
codesona://standards/team): Team-wide rules that apply to all projects
Available MCP Prompts
apply-coding-standards: Applies team coding standards to the conversation via prompt sampling
Troubleshooting
Server won't start:
- Verify
CODESONA_API_KEYis set correctly - Check Node.js version is 18.0.0 or higher
- Ensure API key is valid
Invalid API key:
- Get a fresh API key from Codesona dashboard
- Verify the key has correct permissions
Rules not updating:
- Rules are cached for 5 minutes
- Wait for cache expiry or check Codesona API for updates
IDE can't connect:
- Verify Node.js is in PATH
- Check environment variables are set correctly
- Review IDE logs for detailed error messages
Development
Local Development Setup
- Install dependencies
npm install
- Set environment variables
export CODESONA_API_KEY=your-api-key-here
export CODESONA_API_URL=https://codesona.ai
export NODE_ENV=development
- Run server
node index.js
License
MIT
Support
For issues or questions:
- Visit Codesona documentation
- Open an issue in this repository
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。