AMS Development MCP Servers
MCP servers for real-time development monitoring and interactive pair programming.
README
AMS Development MCP Servers
MCP servers for real-time development monitoring and interactive pair programming
Overview
A suite of Model Context Protocol (MCP) servers designed to enhance AI-assisted development workflows. These servers enable AI assistants like Claude Code to actively monitor development processes, inspect running applications, and participate in interactive debugging sessions.
Packages
This monorepo contains four interconnected MCP servers:
@ams-dev/process-manager · npm
Monitor and control development processes (nx serve, npm run, etc.)
- Start, stop, and restart processes
- Stream and filter process logs in real-time
- Send input to interactive processes
- Pattern-based monitoring for errors and warnings
@ams-dev/docker-manager · npm
Control and inspect Docker containers
- List, start, stop, and restart containers
- Execute commands inside containers
- Stream container logs
- Inspect container configuration and state
@ams-dev/dev-watcher · npm
Real-time error detection and pattern matching across all development sources
- Aggregate logs from processes and containers
- Regex pattern matching for errors and warnings
- Alert deduplication and severity classification
- Integration with browser console logs
@ams-dev/browser-inspector · npm
Browser automation, inspection, monitoring, and performance analysis using Puppeteer
- Navigate and interact with web pages (click, type, wait for elements)
- Monitor console logs and network requests in real-time
- Capture screenshots and extract page content
- Execute custom JavaScript in browser context
- Performance metrics (load times, FCP, resource counts)
- Automated accessibility audits
- Cookie and localStorage management
Quick Start
Installation
# Install all packages globally
npm install -g @ams-dev/process-manager @ams-dev/docker-manager @ams-dev/dev-watcher @ams-dev/browser-inspector
# Or use with npx (no installation required)
npx @ams-dev/process-manager
Configuration
Add to your Claude Code MCP configuration (~/.config/claude/mcp.json):
{
"mcpServers": {
"process-manager": {
"command": "npx",
"args": ["-y", "@ams-dev/process-manager"]
},
"docker-manager": {
"command": "npx",
"args": ["-y", "@ams-dev/docker-manager"]
},
"dev-watcher": {
"command": "npx",
"args": ["-y", "@ams-dev/dev-watcher"]
},
"browser-inspector": {
"command": "npx",
"args": ["-y", "@ams-dev/browser-inspector"]
}
}
}
Use Cases
NX Monorepo Development
Claude: "I'll start the admin and products services for you"
- Starts both services with process-manager
- Monitors for compile errors
- Reports when servers are ready
Docker-based Development
Developer: "Set up the local environment"
Claude: "Starting your database containers"
- Starts PostgreSQL and Redis containers
- Checks container health
- Reports when services are ready
Interactive Debugging
Developer: "Why is the API slow?"
Claude:
- Checks API process logs for slow queries
- Inspects PostgreSQL container logs
- Reports findings with specific line references
Development
This project uses PNPM workspaces for monorepo management.
# Clone the repository
git clone https://github.com/mastoica/vibecoding-mcp-servers.git
cd vibecoding-mcp-servers
# Install dependencies
pnpm install
# Build all packages
pnpm build
# Run tests
pnpm test
Documentation
- Getting Started (coming soon)
- Installation Guide (coming soon)
- Configuration (coming soon)
- Examples (coming soon)
Requirements
- Node.js >= 18.0.0
- PNPM >= 8.0.0
- Docker (optional, for docker-manager)
Contributing
Contributions are welcome! Please read our contributing guidelines before submitting PRs.
License
MIT © mastoica
Roadmap
- [x] Repository setup
- [ ] Process Manager MCP (v0.1.0)
- [ ] Docker Manager MCP (v0.1.0)
- [ ] Dev Watcher MCP (v0.1.0)
- [ ] Complete documentation
- [ ] CI/CD pipeline
- [ ] Publish to npm
Status: 🚧 Work in Progress - Initial development phase
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。