• ANALYSIS
  • REPORTING
  • TRIAGE
  • HUMAN

How to extract Prometheus metrics with your AI

Connect Gatus to your AI client to turn health check data into technical insights. Once connected, your agent can pull specific metrics to investigate outages or performance drifts.

Ask AI about this page

Short answer

How can I pull Gatus metrics into my AI?

Your AI uses the get_metrics tool to pull raw Prometheus-compatible data directly from your Gatus instance. This allows your agent to analyze health check trends and identify the exact cause of failures. You'll get a breakdown of how to set this up and use it below.

Data outcomes.

Where your data goes.

The AI processes raw Gatus metrics into these specific outcomes.

ANALYSIS

Root cause identification

The AI looks at the Prometheus data to find why a health check failed. It identifies patterns in latency or error rates that point to the source of the problem.

REPORTING

Trend summaries

Your agent aggregates metric history to show how service health has changed over time. You get a clear view of stability without manual graphing.

TRIAGE

Automated error grouping

The AI sorts through metric spikes to group similar failure modes. This helps you see if one specific endpoint is causing a cascade of issues.

HUMAN

Manual intervention

The AI flags critical metric thresholds that require you to step in. You decide when to trigger a rollback or a manual service restart.

The workflow

What your AI does when the data arrives.

The AI handles the heavy lifting of data retrieval and interpretation.

  1. Fetch raw data

    Your agent reaches into Gatus to grab the specific Prometheus metrics you need for your query.

    get_metrics

  2. Parse metrics

    The AI translates the raw numerical data into a readable format it can reason about.

    get_metrics

  3. Identify anomalies

    The agent compares current metric values against historical norms to spot deviations.

    get_metrics

  4. Draft findings

    The AI summarizes the technical state of your services based on the retrieved data.

    get_metrics

Try it

Copy these to start.

Use these prompts to begin your investigation.

Starting points

Swap the bracketed text with your actual Gatus endpoint or specific metric names.

Gatus Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_SZx7OyqfGelePhBmZTfpkK6YDYBS4FrcrFgcux0J/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Gatus capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "gatus-health-dashboard-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_SZx7OyqfGelePhBmZTfpkK6YDYBS4FrcrFgcux0J/mcp"
    }
  }
}
Copy into chat04

  • Get the latest Prometheus metrics for the [service name] endpoint in Gatus.

  • Analyze the latency metrics for [endpoint] and tell me if there is a trend.

  • Check the error rate metrics for [service] from the last hour.

  • Summarize the health status of all endpoints in my Gatus instance using the available metrics.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Gatus once, then ask.

Just provide your Gatus credentials once. Your login stays encrypted on our side, and you can immediately start asking your AI about your health checks.

Connect Gatus to your AI

FAQ

How this task behaves.

  • 01

    Can the AI change my Gatus configuration?

    No. The AI can only read metrics through the get_metrics tool. It cannot modify your health checks or settings.

  • 02

    Does the AI delete any of my monitoring data?

    No. The AI has read-only access to the metrics. It cannot delete or alter any data in Gatus.

  • 03

    Can I use this with any MCP-compatible client?

    Yes. Once you connect Gatus via Vinkius, you can use it in Claude, Cursor, Windsurf, or VS Code.

  • 04

    What kind of data can the AI see?

    The AI sees any Prometheus-compatible metric exposed by your Gatus instance.

  • 05

    Do I need to install anything on my local machine?

    No. Vinkius hosts the MCP, so you only need to connect your AI client.

  • More questions about Gatus? The Connector page answers them. See everything the Gatus Connector can do

Connect Gatus to Claude, Cursor, ChatGPT & more