vcenter-mcp
Enables plain English queries about VMware vCenter environments including clusters, VMs, datastores, and more, directly from VS Code via GitHub Copilot.
README
vcenter-mcp
Ask GitHub Copilot questions about your Broadcom VCF and VMware vCenter environment in plain English — clusters, datacenters, datastores, folders, hosts, networks, resource pools, VMs, guest details, appliance version and time, and more — directly from VS Code.
Supports Broadcom VCF vCenter Server Appliance and VMware vCenter Server Appliance — with per-vCenter credentials stored securely in your OS keyring (Windows Credential Manager / macOS Keychain / Linux Secret Service).
Requirements
- Python 3.10+
- VS Code
- GitHub Copilot + Copilot Chat extensions
Quick start
# 1. Install
pip install vcenter-mcp
# 2. Create a folder for your inventory, navigate to it, and run the wizard
mkdir my-vcenter && cd my-vcenter
vcenter-mcp configure
The wizard will:
- Verify Python 3.10+ is active
- Install / update all dependencies
- Register your vCenter Server Appliance instance(s) in
inventory.yaml - Store credentials securely in the OS keyring — no plain-text passwords anywhere
- Write
.vscode/mcp.jsonpointing VS Code at the server (no credentials in this file)
# 3. Open the folder in VS Code
code .
In VS Code, open Copilot Chat, switch to Agent mode, and start asking questions.
Already set up? Re-run
vcenter-mcp configureat any time to add new vCenter instances and to update credentials.
Configuration
Credential storage
Credentials are stored in the OS keyring under the service name vcenter-mcp. Key format:
| Entity | Keyring key examples |
|---|---|
| vCenter (default) | vcenter.default.username, vcenter.default.password |
| vCenter (override) | vcenter.VCENTER-NAME.username, vcenter.VCENTER-NAME.password |
Lookup falls back from named entry → vcenter.default.* → environment variables (VCENTER_USERNAME etc.) for compatibility and automation.
vCenter inventory — inventory.yaml
Register your instances here. Edit directly or re-run vcenter-mcp configure.
vcenters:
- name: PROD-VCENTER-1
fqdn: vcsa01.example.local
ip_address: 192.168.1.10
verify_ssl: false
- name: PROD-VCENTER-2
fqdn: vcsa02.example.local
ip_address: 192.168.1.11
verify_ssl: true
name— label used in prompts and tool parameters (vcenter_name)fqdnorip_address— at least one is required; if both are provided,fqdnis preferred- Single entry → selected automatically; multiple entries → specify the name in your prompt
VS Code — .vscode/mcp.json
Generated by vcenter-mcp configure. Contains no credentials:
{
"servers": {
"vcenter-mcp": {
"type": "stdio",
"command": "vcenter-mcp",
"env": {
"VCENTER_MCP_INVENTORY": "${workspaceFolder}/inventory.yaml"
}
}
}
}
Usage
Switch Copilot Chat to Agent mode and ask in plain English:
List all VMs and their power state
Show me all clusters and datacenters
Which hosts are in the cluster and what version are they running?
List all datastores and their free capacity
Show me the networks in vCenter
Get guest identity details for VM <vm-id>
Show me the appliance version and time
Target a specific vCenter by name:
List all VMs from PROD-VCENTER-1
Show appliance version for PROD-VCENTER-1
List all datastores from PROD-VCENTER-2
Copilot calls list_vcenters automatically when needed to discover available entries.
Project structure
vcenter-mcp/
├── inventory.yaml vCenter registry (user-generated)
├── pyproject.toml Package metadata and dependencies
├── available_tools.md Full tool reference
├── SECURITY_POSTURE.md Security posture, cautions, and operational guidance
└── src/
└── vcenter_mcp/
├── __init__.py
├── app.py Shared FastMCP instance ("vcenter-mcp")
├── cli.py Entry point — 'vcenter-mcp' and 'vcenter-mcp configure'
├── server.py Thin wrapper delegating to cli.main()
├── credentials.py OS keyring read/write helpers
├── registry.py Inventory loading, host+credential resolver, JSON helper
├── client.py HTTP client: session auth and policy-validated GET requests
├── security.py Central security policy, limits, and startup config validation
├── vsphere_ws.py vSphere Web Services helpers for ESXi host detail
└── tools/
├── inventory.py list_vcenters, get_vcenter_inventory
├── vcenter_inventory.py list_clusters, get_cluster,
│ get_cluster_resource_utilization_ws,
│ get_cluster_cpu_memory_utilization_period_ws,
│ get_cluster_cpu_memory_daily_rollup_ws,
│ get_cluster_cpu_memory_utilization_window_ws,
│ list_datacenters, get_datacenter,
│ list_datastores, get_datastore,
│ list_folders,
│ list_hosts, get_host,
│ list_networks,
│ list_resource_pools, get_resource_pool,
│ list_vms, get_vm
├── vm_details.py get_vm_guest_identity, list_vm_guest_local_filesystems,
│ list_vm_guest_network_interfaces, get_vm_hardware,
│ get_vm_boot, get_vm_cpu, get_vm_memory,
│ list_vm_disks, get_vm_disk,
│ list_vm_nics, get_vm_nic
└── appliance.py get_appliance_version, get_appliance_time
See available_tools.md for the full tool reference including parameters, return values, and source modules.
Notes
- All tools are read-only — no changes are made to your vCenter environment
- Uses the vCenter REST API at
/api/*over HTTPS, typically port 443 - Uses the vSphere Web Services API through vCenter for
list_hostsandget_host, because REST host detail coverage is not available in vCenter API surface - Cluster window utilization tool behavior: supports
1-30day windows and hour inputs only when they are multiples of 24 (for example24 hours,168 hours) - Historical utilization output depends on PerformanceManager retention in your vCenter; requested windows may return fewer samples if older data is not retained
- Authentication uses
POST /api/sessionand reuses the session token for tool calls - If multiple vCenters are configured, specify
vcenter_nameto target the correct instance - Transport is stdio — the server runs locally and is managed by VS Code
Security
See SECURITY_POSTURE.md for current security controls, hardening behavior, and operational cautions.
Disclaimer
This is an independent, community-built project and is not an official Broadcom or VMware product. It is not affiliated with, endorsed by, or supported by Broadcom Inc. or VMware by Broadcom in any way. VMware, vSphere, vCenter, and VCF are trademarks of their respective owners.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.