# Checkly MCP for AI Agents AI Agent Connect

> Checkly MCP lets you manage your application monitoring and synthetic testing through your AI client. You can track API uptime, trigger manual checks, and audit your alert channels without leaving your chat interface. It connects your monitoring stack directly to your workspace for faster responses to system issues and better visibility into your backend health.

## Overview
- **Category:** loved-by-devs
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_ZKVvNJK1bEgGdJP2vFYcWaZwHDzM0XtD3oz2UTx3/ai-agent-connect
- **Tags:** api-monitoring, synthetic-testing, uptime-tracking, end-to-end-testing, performance-monitoring

## Description

You're tired of jumping between tabs to see if your API is actually up or if your cron jobs are firing correctly. This Connector changes that by bringing your entire monitoring stack into your chat window. Instead of hunting through a dashboard to find a specific monitor's performance metrics, you just ask your agent to pull them. If you need to run a manual check to see how a specific endpoint handles a spike, you can do that instantly. You can also audit where your alerts are actually going, whether they're hitting Slack or PagerDuty, to make sure your team isn't missing critical updates. It's about getting the data you need to keep your systems healthy without the overhead of manual navigation. By connecting this to the Vinkius catalog, you give your agent the eyes it needs to oversee your infrastructure in real time.

## Tools

### get_checkly_account_info
Pull core account and organization metadata to understand your workspace structure. Use this to see who has access and how your account is organized.

### get_check_performance_metrics
Get specific performance data and response times for any given check. This helps you identify exactly when and where your application started slowing down.

### get_check_details
Retrieve full details for a specific monitor to see its current configuration. It's the fastest way to see the exact parameters of a specific test.

### list_checkly_alert_channels
See all configured alert destinations like Slack, Email, or PagerDuty. This ensures your team knows exactly where they'll receive critical notifications.

### list_checkly_checks
Get a full list of all your active API and Browser monitors. Use this to get a quick overview of everything you're currently tracking.

### list_check_groups
View the different groups your checks are organized into. This helps you see how your monitoring is structured across different services.

### list_checkly_heartbeats
List all heartbeat and cron monitors to ensure background tasks are running. It's a great way to verify that your nightly jobs are still active.

### trigger_check_run
Manually start a check run immediately to verify system health on demand. This is useful when you need to see how a system handles a specific load right now.

## Prompt Examples

**Prompt:** 
```
Are my monitors up?
```

**Response:** 
```
### Current Monitor Status

| Monitor Name | Status | Last Check |
| :--- | :--- | :--- |
| **Main API** | ✅ Passing | 2 mins ago |
| **Checkout Flow** | ✅ Passing | 5 mins ago |
| **Auth Service** | ❌ Failing | 1 min ago |
| **DB Cleanup** | ✅ Active | 45 mins ago |

Would you like me to trigger a manual check on the Auth Service?
```

**Prompt:** 
```
Show me the response time for the checkout flow.
```

**Response:** 
```
### Performance Metrics: Checkout Flow

* **Average Response Time:** 450ms
* **Peak Response Time:** 1.2s (at 3:00 AM)
* **95th Percentile:** 680ms
* **Uptime:** 100%

*Note: These metrics cover the last 24 hours.*
```

**Prompt:** 
```
Who is getting alerted if the API fails?
```

**Response:** 
```
### Configured Alert Channels

Your alerts are currently routed to the following destinations:

* **Slack:** `#ops-alerts-prod`
* **Email:** `oncall-team@company.com`
* **PagerDuty:** `Service: API-Gateway-Production`

All active monitors are currently linked to these channels.
```

## Capabilities

### Trigger immediate check runs
Manually start a check to verify system health on demand.

### Pull performance metrics for specific monitors
Get specific performance data and response times for any given check.

### List all API and browser checks
Get a full list of all your active API and Browser monitors.

### Audit alert channels like Slack and PagerDuty
See all configured alert destinations to ensure your team gets notified.

### Monitor heartbeat and cron job status
List all heartbeat and cron monitors to ensure background tasks are running.

### Retrieve organization and account metadata
Pull core account and organization metadata to understand your workspace structure.

## Use Cases

### Quick health check
An engineer sees a spike in errors and asks the agent to trigger_check_run on the auth service to see if it's a transient blip.

### Audit review
An SRE needs to verify if the production alerts are hitting the right Slack channel and uses list_checkly_alert_channels.

### Performance deep dive
A developer wants to know why a specific API is slow and asks for the metrics from get_check_performance_metrics.

### Cron job verification
A backend dev wants to ensure the nightly cleanup is still running and asks the agent to list_checkly_heartbeats.

## Benefits

- Stop switching tabs by pulling monitor details and performance metrics directly into your chat interface.
- Run manual checks on demand using trigger_check_run to verify system health without navigating a dashboard.
- Audit your entire notification stack using list_checkly_alert_channels to ensure your team gets the right pings.
- Keep track of background tasks by using list_checkly_heartbeats to confirm cron jobs are firing as expected.
- Get a bird's-eye view of your infrastructure by pulling all monitors with list_checkly_checks in one command.
- Access organizational metadata with get_checkly_account_info to manage permissions and structures more easily.

## How It Works

The bottom line is you get a conversational interface for your entire monitoring and synthetic testing stack.

1. Subscribe to the Checkly MCP on Vinkius.
2. Plug in your Checkly API Key and Account ID from your settings.
3. Ask your AI client to list monitors or trigger a check run.

## Frequently Asked Questions

**What can I do with the Checkly MCP?**
You can use it to manage your monitoring stack through your AI client. This includes checking API uptime, pulling performance metrics, and auditing your alert configurations without leaving your chat interface.

**Can I trigger manual checks with Checkly MCP?**
Yes, you can ask your agent to run a check immediately. This is great for verifying system health during a deployment or after a hotfix.

**How do I see my alert channels using Checkly MCP?**
Just ask your agent to list your alert channels. It will pull all your configured destinations, such as Slack, Email, and PagerDuty, into one view.

**Can Checkly MCP monitor my cron jobs?**
Yes, it can. You can use it to list and monitor your heartbeat and cron jobs to ensure your background tasks are running as they should.

**Does Checkly MCP show performance data?**
It does. You can request specific performance metrics for any monitor to see response times and other key health data.

**How do I connect Checkly MCP to my AI client?**
Once you subscribe on Vinkius, you just need to provide your Checkly API Key and Account ID. Your agent will then be able to interact with your monitoring stack.

**Can I trigger an API check manually through the agent?**
Yes! Use the `trigger_check_run` tool with the unique check ID. The agent will instruct Checkly to run the test immediately from its global network.

**How do I see the performance metrics for my homepage check?**
Use the `get_check_performance_metrics` tool with the check ID. Your agent will fetch detailed analytics including success rates and response time percentiles.

**Where do I find my Checkly API Key and Account ID?**
Log in to your Checkly account and navigate to **User Settings -> API Keys** for the key. The **Account ID** can be found in the **Account Settings** section.