• AUDIT
  • SETUP
  • ALERT
  • HUMAN

How to manage AI performance monitors with your AI

Connect Datadog to your agent to watch LLM health in real time. Once linked, your AI handles the heavy lifting of tracking service thresholds and response quality.

Ask AI about this page

Short answer

How can I manage AI performance monitors with my AI?

You tell your agent to check your existing Datadog monitors or instruct it to build new ones for specific LLM metrics. It pulls live data and sets up the triggers you need to catch service degradation immediately.

Monitoring outcomes.

Where your observability data lands.

The AI organizes your Datadog tasks into these four areas.

AUDIT

Existing monitor review

The AI pulls your current Datadog setup so you can see exactly what is already being tracked.

SETUP

New threshold creation

Your agent builds fresh monitors to watch for specific LLM response failures or latency spikes.

ALERT

Health status updates

The AI monitors the health of your LLM services and flags when performance dips below your set limits.

HUMAN

Manual intervention

You decide when to adjust thresholds or delete old monitors based on the AI's findings.

The workflow

What your AI does when <the thing arrives>.

The AI acts as your observability operator, moving between checking status and setting new rules.

  1. Inventory check

    The agent scans your Datadog account to find every active AI performance monitor currently running.

    list_ai_monitors

  2. Threshold definition

    You describe the specific error or latency limit you want to watch, and the AI builds the monitor.

    create_monitor

  3. Health verification

    The AI checks the current status of your LLM response health against the monitors it just listed.

    list_ai_monitors

  4. Rule updates

    The agent creates new monitoring rules to cover new service endpoints or updated model versions.

    create_monitor

Try it

Copy these to start.

Use these prompts to kick off your monitoring tasks.

Starting points

These are basic requests. Swap out the specific monitor names or metric types to match your Datadog setup.

Datadog Connector

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

Connector linkhttps://edge.vinkius.com/vk_preview_bWdhwHEHDioKsLvhV1Lvv7ec4t2gm8k5cIgCZeWx/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 — Datadog capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "datadog-ai-llm-observability-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_bWdhwHEHDioKsLvhV1Lvv7ec4t2gm8k5cIgCZeWx/mcp"
    }
  }
}
Copy into chat04

  • List all my current AI performance monitors in Datadog.

  • Create a new monitor for LLM latency that triggers if response time exceeds 2 seconds.

  • Check my existing monitors to see if any LLM health thresholds are currently being breached.

  • Set up a new monitor to track error rates for my production LLM service.

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

Start here

Connect Datadog AI once, then ask.

Just link your Datadog account through the Vinkius interface. Your credentials stay encrypted, and you can immediately start managing monitors through your preferred AI client.

Connect Datadog AI to your AI

FAQ

How this task behaves.

  • 01

    Can the AI delete my existing Datadog monitors?

    No. The AI can only list and create monitors. You must manually delete any monitors in Datadog if they are no longer needed.

  • 02

    Can the AI change the threshold of a monitor I already made?

    No. The AI can create new monitors with different thresholds, but it cannot modify the settings of an existing monitor.

  • 03

    Does the AI need my Datadog API key?

    Yes, you provide your credentials during the connection process, but Vinkius keeps them encrypted and secure.

  • 04

    Can I use this with Claude or Cursor?

    Yes. Any MCP-compatible client like Claude, Cursor, or Windsurf can use this connector once it is connected.

  • 05

    How does the AI know which monitors to look at?

    The AI uses the list tool to pull all available AI-related monitors from your Datadog account to identify them.

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

Connect Datadog to Claude, Cursor, ChatGPT & more