Auralis Commander
A lightweight Windows-native MCP server providing a consolidated suite of 14 tools for shell execution, file operations, and interactive process management. It optimizes efficiency through batch file operations and smart process handling to minimize context window overhead.
README
Auralis Commander
The lightweight Windows MCP server that does more with less.
Why Auralis Commander?
We built Auralis Commander because existing solutions were either too heavy or too limited:
| MCP Server | Tools | Focus | Problem |
|---|---|---|---|
| Desktop Commander | 26 | Everything | Bloated, 5 tools just for processes |
| Filesystem (Anthropic) | 11 | Files only | No shell, no processes, no system info |
| Windows CLI | 8 | PowerShell | No file operations, limited scope |
Auralis Commander: 14 tools that cover shell, files, search, processes, and system — without the bloat.
Key Advantages
🎯 Smart Design
One process_interactive tool replaces 5 separate tools in Desktop Commander:
start_process→process_interactive { action: "start" }read_process_output→process_interactive { action: "read" }interact_with_process→process_interactive { action: "write" }force_terminate→process_interactive { action: "kill" }list_sessions→process_interactive { action: "list" }
📦 Batch Operations
Read multiple files in one call:
file_read { paths: ["config.json", "package.json", ".env"] }
// Returns all files at once, errors don't block other files
⚡ Windows-Native
Optimized for PowerShell and Windows workflows. No WSL required, no Unix assumptions.
🪶 Lightweight
~14KB of focused code. Fast startup, minimal memory footprint, smaller context window usage.
Tool Comparison
| Capability | Auralis | Desktop Commander | Filesystem | Windows CLI |
|---|---|---|---|---|
| Shell execution | ✅ | ✅ | ❌ | ✅ |
| File read/write | ✅ | ✅ | ✅ | ❌ |
| File search | ✅ | ✅ | ✅ | ❌ |
| Batch file read | ✅ | ✅ | ✅ | ❌ |
| Interactive processes | ✅ | ✅ | ❌ | ❌ |
| Process management | ✅ | ✅ | ❌ | ❌ |
| System info | ✅ | ✅ | ❌ | ✅ |
| Find & replace | ✅ | ✅ | ✅ | ❌ |
| Total tools | 14 | 26 | 11 | 8 |
| Context overhead | Low | High | Low | Low |
Installation
Option 1: npm (recommended)
npm install -g auralis-commander
Option 2: Clone and Build
git clone https://github.com/antonpme/auralis-commander
cd auralis-commander
npm install
npm run build
Claude Desktop Configuration
Add to your claude_desktop_config.json:
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"auralis-commander": {
"command": "node",
"args": ["C:/path/to/auralis-commander/dist/index.js"]
}
}
}
Restart Claude Desktop to load the server.
Tools Reference
Shell & System
| Tool | Description |
|---|---|
shell_exec |
Execute commands in pwsh, powershell, or cmd |
system_info |
Get CPU, memory, disk usage, and uptime |
processes |
List running processes with memory/CPU stats |
process_kill |
Terminate process by PID or name |
process_interactive |
Run and interact with long-running processes |
Files
| Tool | Description |
|---|---|
file_read |
Read single file or batch with paths array |
file_write |
Write or append to file, auto-create directories |
file_edit |
Find and replace text in files |
file_delete |
Delete files or directories (with recursive flag) |
file_move |
Move or rename files and directories |
file_info |
Get metadata: size, dates, line count, permissions |
Directories & Search
| Tool | Description |
|---|---|
dir_list |
List contents with depth control and glob patterns |
dir_create |
Create directory with parent directories |
search |
Search by filename or content with filtering |
Usage Examples
Run a Dev Server
// Start Next.js dev server
process_interactive {
action: "start",
command: "npm run dev",
cwd: "C:/projects/my-app"
}
// → { session_id: "a1b2c3", output: "ready on http://localhost:3000", is_running: true }
// Check for new output
process_interactive { action: "read", session_id: "a1b2c3", timeout_ms: 5000 }
// Stop when done
process_interactive { action: "kill", session_id: "a1b2c3" }
Interactive Python REPL
process_interactive { action: "start", command: "python -u -i" } // -u: unbuffered, -i: interactive
// → { session_id: "x1y2z3", output: "Python 3.12.0\n>>>", is_running: true }
process_interactive { action: "write", session_id: "x1y2z3", input: "2 + 2\n" }
// → { output: "4\n>>>", is_running: true }
process_interactive { action: "write", session_id: "x1y2z3", input: "exit()\n" }
// → { output: "", is_running: false }
Batch Configuration Check
file_read {
paths: [
"package.json",
"tsconfig.json",
".env",
".env.local"
]
}
// Returns all files; missing ones show error without blocking others
Quick System Health Check
system_info {}
// → { cpu: { model, cores, usage }, memory: { total, used, free }, disks: [...], uptime: "3d 14h" }
processes { sort_by: "memory", limit: 10 }
// → Top 10 memory consumers
Configuration
Create auralis-commander.json in your home directory for defaults:
{
"default_shell": "pwsh",
"default_cwd": "C:/Projects",
"max_file_read_mb": 50
}
| Option | Default | Description |
|---|---|---|
default_shell |
pwsh |
Shell for commands: pwsh, powershell, or cmd |
default_cwd |
Home dir | Default working directory |
max_file_read_mb |
50 |
Maximum file size for reading |
Architecture
auralis-commander/
├── src/
│ ├── index.ts # MCP server setup & tool registration
│ ├── config.ts # Configuration management
│ ├── tools/
│ │ ├── shell.ts # shell_exec
│ │ ├── files.ts # file_* and dir_* operations
│ │ ├── search.ts # Content and filename search
│ │ ├── processes.ts # Process listing and killing
│ │ ├── system.ts # System information
│ │ └── interactive.ts # Interactive process sessions
│ └── utils/
│ ├── powershell.ts # PowerShell execution wrapper
│ ├── paths.ts # Path normalization
│ └── errors.ts # Error handling
├── dist/ # Compiled JavaScript
└── package.json
Contributing
Issues and PRs welcome. Please:
- Keep tools focused — no feature creep
- Maintain Windows compatibility
- Test with Claude Desktop before submitting
License
MIT License — use it, modify it, ship it.
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。