# Checkmk MCP for AI Agents MCP

> Checkmk lets you manage your IT infrastructure monitoring through natural conversation. You can query service states, list host details, check configuration groups, or even activate pending changes—all directly from your AI client without opening the complex GUI.

## Overview
- **Category:** cloud-infrastructure
- **Price:** Free
- **Tags:** server-monitoring, infrastructure-management, network-monitoring, it-operations, real-time-metrics

## Description

Managing a large server landscape usually means clicking through dashboards and running specific reports in a dedicated console. With this MCP, you talk to your agent instead. You get full control over your entire IT monitoring setup by asking questions like, 'What's wrong with the web servers?' or 'Show me all services on the database cluster.'

Your AI client handles the complex calls, giving you immediate status reports and diagnostic data right in your chat window. Need to audit which hosts belong to a specific group? Just ask. Want to push through configuration changes that are waiting for approval? It’s one prompt away.

It takes the headache out of maintaining service visibility. Because this MCP is hosted on Vinkius, you connect once and gain access not just to Checkmk controls, but to thousands of other infrastructure tools your team needs.

## Tools

### activate_checkmk_changes
Activates configuration changes that are currently pending in the Checkmk monitoring setup.

### get_host_details
Retrieves detailed, technical information for any specific host you name.

### list_all_monitored_services
Provides a complete list of all services running across every monitored host.

### list_checkmk_folders
Lists all major configuration folders used to organize the monitoring structure.

### list_host_groups
Retrieves a list of configured groups that contain multiple hosts.

### list_checkmk_hosts
Lists every individual host currently under monitoring.

### list_service_groups
Provides a list of service groups, helping you understand how services are organized.

### list_host_services
Retrieves all monitored services associated with one specific host.

## Prompt Examples

**Prompt:** 
```
List all my monitored hosts and tell me which ones are currently down.
```

**Response:** 
```
**🌐 Host Status Report**

*   `web-prod-01`: **UP** (Service: OK)
*   `db-primary`: **UP** (Service: WARNING - Disk Space)
*   `api-gateway`: **DOWN** 🔴 (Status: Critical failure detected. Needs immediate attention.)
*   `auth-replica`: **UP** (Service: OK)

I found 12 total hosts. `api-gateway` is the only one flagged as down.
```

**Prompt:** 
```
Show me all services running on 'web-prod-01' and tell me which ones are not OK.
```

**Response:** 
```
🔍 **Service Report for web-prod-01**

| Service Name | Status | Metric Value |
| :--- | :---: | :---: |
| CPU Load | WARNING | 4.5% |
| Filesystem /var | CRITICAL | 98% full |
| Network Latency | OK | 12ms |

**Action Required:** The filesystem on `/var` is critical. You should check disk usage immediately.
```

**Prompt:** 
```
I need to see the details for the host 'db-primary' and list all services it runs.
```

**Response:** 
```
**🤖 Detailed Host Report: db-primary**

*   **Host ID:** 3a2b9c
*   **Owner Group:** Database Team (ID: 5)
*   **Uptime:** 14 days, 3 hours

**Associated Services:**
*   PostgreSQL Replication (OK)
*   Backup Agent (WARNING - Last run failed)
*   Monitoring Agent (OK)
```

## Capabilities

### Check all monitored hosts
Retrieve a comprehensive list of every host currently being tracked by the system.

### Diagnose service health on specific machines
Get real-time status and detailed metrics for services running on any given host.

### Pull detailed machine configurations
Fetch in-depth, technical information about a single monitored server or appliance.

### Review the entire monitoring structure
List and map out all configured folders, host groups, and service groups to understand your setup's architecture.

### Activate pending changes
Manually trigger the application of configuration updates across the monitored infrastructure.

## Use Cases

### Auditing service status across the environment
The ops analyst needs to verify if a critical application is running on every server. Instead of checking dozens of individual dashboards, they ask their agent to run `list_all_monitored_services`, getting one consolidated report showing service health everywhere.

### Preparing for maintenance window changes
A DevOps engineer needs to push a new monitoring configuration. They use the MCP to list all relevant host groups (`list_host_groups`) and then trigger `activate_checkmk_changes` right from their agent, confirming deployment readiness.

### Investigating an intermittent server failure
A support team member hears about a slow machine. They use the MCP to first run `list_checkmk_hosts` to confirm the host is tracked, then use `get_host_details` to check core system metrics and plugin outputs for immediate clues.

### Mapping out an unfamiliar monitoring setup
A new team member needs to understand the organizational structure. They ask their agent to call `list_checkmk_folders`, followed by checking service groupings (`list_service_groups`), building a full architectural map via conversation.

## Benefits

- Don't open the GUI to check status. You ask your agent, and it runs commands like `list_checkmk_hosts` instantly, providing a list of all monitored servers.
- When troubleshooting, you don't waste time digging for data. Use the agent to run diagnostics by listing services on a host using `list_host_services`, getting immediate issue details.
- Need an audit trail? The MCP lets you map out your entire setup structure by calling functions like `list_checkmk_folders` and `list_service_groups` without navigating menus.
- Configuration changes are simple. Instead of a multi-step process, just ask the agent to activate pending changes using `activate_checkmk_changes` and watch it run.
- Get full context on any machine by running `get_host_details`. This pulls all core system metadata you need in one chat response.

## How It Works

The bottom line is you stop navigating complex GUIs and start talking to your infrastructure monitoring system instead.

1. First, subscribe to this MCP and provide your Checkmk credentials, including the Username, Automation Secret, and Server URL.
2. Next, connect your preferred AI agent client. Your agent now has direct access to all of Checkmk's monitoring APIs through Vinkius.
3. Finally, simply ask a natural language question—like 'List all hosts that are critical'—and the agent performs the necessary checks and delivers an actionable report.

## Frequently Asked Questions

**How can Checkmk MCP help me monitor my servers without using the web interface?**
You can manage everything conversationally. Instead of clicking around dashboards, you ask your agent to list hosts (`list_checkmk_hosts`) or check service groups (`list_service_groups`). It gives you real-time metrics and reports directly in the chat.

**Can I use Checkmk MCP to activate changes on my infrastructure?**
Yes, this is one of its key features. You simply ask your agent to trigger pending configuration updates using `activate_checkmk_changes`. It handles the activation run and reports success or failure immediately.

**What if I need detailed info on just one server?**
You don't have to guess. You can ask for specific host details using `get_host_details` for a single machine, pulling all core system metadata and configuration data in a structured report.

**Is Checkmk MCP useful for auditing my monitoring setup?**
Absolutely. The MCP lets you list the entire structure—like finding out every folder (`list_checkmk_folders`) or service group (`list_service_groups`)—giving you a full audit trail of your infrastructure organization.

**Does Checkmk MCP help me troubleshoot live issues?**
Yes. When an issue pops up, you can ask the agent to run diagnostics by listing services on that specific host (`list_host_services`), which retrieves plugin output and current states for immediate troubleshooting.