vme-mcp-cli 0.1.3

VME Infrastructure Management CLI Tool - Chat interface with voice commands

VME/Morpheus FastMCP Server v2

Advanced MCP server for HPE VM Essentials (VME) and Morpheus infrastructure management with automatic platform detection and intelligent tool filtering.

NEW: Now includes a sophisticated VME Textual CLI Client with OpenAI Realtime API audio integration for voice-enabled infrastructure management.

Features

• 🔍 Automatic Platform Detection - Detects VME vs Morpheus Enterprise at startup
• ⚡ Dynamic Route Filtering - Only exposes available endpoints based on platform capabilities
• 🛠️ Comprehensive Tool Suite - 17+ VME management tools for VMs, storage, networking
• 🎙️ Voice-Enabled Chat Interface - OpenAI Realtime API integration with microphone support
• 🧪 Advanced Testing Framework - Multi-test suite with single test execution
• 🏗️ Real Infrastructure Integration - Works with live VME/Morpheus deployments

Quick Start

🚀 VME Textual CLI Client (NEW)

Manage VME infrastructure through natural language chat with voice support:

Audio Setup (For Microphone Support)

Fedora/RHEL:

# Install PortAudio for real microphone access
sudo dnf install -y portaudio-devel

# Install Python audio dependencies
uv add pyaudio

# Verify audio system
pactl info

Ubuntu/Debian:

sudo apt-get install portaudio19-dev
uv add pyaudio

Quick Start

1. Setup Environment

   # Edit .env with your credentials
   echo "VME_API_BASE_URL=https://your-vme-server/api" >> .env
   echo "VME_API_TOKEN=your-api-token" >> .env
   echo "ANTHROPIC_API_KEY=your-anthropic-key" >> .env
   echo "OPENAI_API_KEY=your-openai-key" >> .env  # For audio features
   

2. Install Dependencies

   uv sync
   

3. Launch Client

   # Simple launch (loads API keys from .env)
   python launch_client.py
   
   # Or use module directly
   python -m src.clients.textual_cli.main
   

4. Configure Audio (Optional)

   # Edit vme_client.yaml to enable voice features
   audio:
     enabled: true                    # Enable audio support
     voice: "alloy"                   # OpenAI voice (alloy, echo, fable, onyx, nova, shimmer)
     model: "gpt-4o-realtime-preview-2024-10-01"
   

5. Start Chatting

   💬 Try these commands (type or speak):
   • "What tools are available?"
   • "Discover compute infrastructure"  
   • "List all VMs"
   • "Get appliance settings"
   • "Create a VM called test-server"
   

   Audio Features:

   • 🎤 Microphone Status: Visual indicator when listening/speaking
   • 🔊 Voice Responses: AI responds with speech + text
   • 📊 Audio Levels: Real-time visualization of microphone input
   • 🎛️ System Integration: Linux system tray shows mic activity

🔧 MCP Server Only

1. Setup Environment

   # Copy environment template
   cp .env.example .env
   
   # Edit with your VME/Morpheus credentials
   vim .env
   

2. Install Dependencies

   uv sync
   

3. Run Server

   # Auto-detecting server (recommended)
   uv run python vme_server_filtered.py
   
   # With MCP Inspector for testing
   ./scripts/test-with-inspector.sh
   

Testing

# Run all tests
uv run python tests/test_fastmcp_client.py

# Run single test
uv run python tests/test_fastmcp_client.py check_license_info

# Available test cases
uv run python tests/test_fastmcp_client.py invalid_name

VME Textual CLI Client

Features

🎨 Rich Terminal Interface

• GitHub-inspired dark theme with professional styling
• Real-time thinking indicators with animations
• Markdown rendering for LLM responses
• Clean chat interface with syntax highlighting

🧠 AI-Powered Infrastructure Management

• Natural language VM creation and management
• Intelligent tool discovery and calling
• Context-aware infrastructure operations
• Support for both Anthropic Claude and OpenAI GPT

🔧 Progressive Discovery Workflow

• Start with 18 basic tools
• Unlock tool groups on demand: discover_compute_infrastructure() → 40+ VM tools
• Logical grouping: compute, networking, storage, monitoring, management
• Guided workflows for complex operations

⚙️ Smart Configuration

• YAML-based configuration with environment variable resolution
• Auto-loading from .env files
• Configurable timeouts, themes, and LLM settings
• Validation for API keys and server connectivity

Client Architecture

src/clients/textual_cli/
├── main.py              # CLI entry point with Click interface
├── config/
│   └── settings.py      # YAML config system with .env support
├── ui/
│   └── app.py          # Textual interface (GitHub theme, chat area)
├── mcp/
│   └── manager.py      # FastMCP client integration
└── llm/
    └── manager.py      # LLM providers (Anthropic/OpenAI)

launch_client.py         # Simple launcher script
vme_client.yaml         # Default client configuration
test_client.py          # Component testing utility

Usage Examples

# Basic infrastructure queries
"What tools are available?"
"Get appliance settings"
"Show current license information"

# VM Management  
"List all virtual machines"
"Create a VM called 'web-server-01' with large size"
"Show me VM creation options"

# Progressive Discovery
"Discover compute infrastructure"     # Unlocks 40+ VM tools
"Discover networking capabilities"    # Unlocks network tools  
"Discover storage capabilities"       # Unlocks storage tools
"What infrastructure is available?"   # Overview of all capabilities

# Advanced Operations
"Create a web server VM with 4 CPUs"
"Show me all running VMs in zone us-east-1"
"What instance types are available?"

Configuration

Default config (vme_client.yaml):

llm:
  anthropic_key: ${ANTHROPIC_API_KEY}
  default_provider: anthropic
  default_model: claude-3-5-sonnet-20241022

server:
  script_path: src/servers/progressive_discovery_server.py
  auto_connect: true
  timeout: 30

ui:
  theme: github_dark
  show_thinking_indicator: true
  max_message_history: 1000

Architecture

   api_utils.py              # Shared API utilities and platform detection
   vme_server_filtered.py    # Main MCP server with filtering
   src/servers/
      progressive_discovery_server.py  # Progressive discovery for clients
      vme_server_adaptive.py          # Client-adaptive server  
   src/clients/textual_cli/           # Textual CLI client
   tests/
      test_fastmcp_client.py # Comprehensive test suite
   config/                   # YAML route configurations (future)
   scripts/                  # Testing and development scripts

Platform Detection

The server automatically detects platform type using license endpoint:

from api_utils import detect_platform_type

platform = detect_platform_type(api_url, token)
# Returns: 'vme', 'morpheus', or 'unknown'

Available Tools

VME Core Tools

• VM Management: Create, list, manage instances
• Platform Info: License status, appliance settings
• Instance Types: Provisioning options and configurations
• Infrastructure: Basic storage, networking, zones

Enterprise Tools (Morpheus Only) 🏢

• Advanced reporting and analytics
• Workflow automation
• Multi-cloud management
• Enterprise billing and costing

Configuration

Environment Variables

VME_API_BASE_URL=https://your-vme-server/api
VME_API_TOKEN=your-api-token
ANTHROPIC_API_KEY=your-anthropic-key  # For testing

MCP Client Configuration

{
  "mcpServers": {
    "vme-infrastructure": {
      "command": "uv",
      "args": ["run", "python", "vme_server_filtered.py"],
      "env": {
        "VME_API_BASE_URL": "https://vmemgr01.lab.loc/api",
        "VME_API_TOKEN": "your-token-here"
      }
    }
  }
}

Development

1. Platform Detection Testing

   uv run python tests/test_fastmcp_client.py check_license_info
   

2. VM Operations Testing

   uv run python tests/test_fastmcp_client.py list_running_vms
   

3. Full Test Suite

   uv run python tests/test_fastmcp_client.py
   

Known Issues

• Claude Code MCP client has bugs with tool name limits (use MCP Inspector for testing)
• VME returns 403 for enterprise-only endpoints (expected behavior)
• FastMCP route filtering logs may interfere with stdio transport

License

Private repository - Internal use only.