- ALERT LIST
- NODE DATA
- ROOT CAUSE
- HUMAN ACTION
Short answer
How can I use AI to triage Netdata alarms?
Your AI pulls a list of active alerts from your Netdata Cloud spaces and fetches specific chart data for the affected nodes. This lets you jump straight from a notification to a deep dive into the underlying metrics.
Investigation outcomes
Where your investigation lands.
The AI organizes your telemetry into actionable insights.
ALERT LIST
Identify active issues
The AI scans your Netdata Cloud spaces to find every running alarm. You get a clear view of what's broken right now.
NODE DATA
Inspect specific nodes
Once you find a problem, the AI pulls the exact metrics for that node. You see the specific spikes or drops causing the alert.
ROOT CAUSE
Pinpoint the failure
The AI compares different metrics to see if a CPU spike caused a memory leak or a disk issue. It connects the dots between different charts.
HUMAN ACTION
Decide the fix
The AI presents the evidence, but you make the call. You decide whether to restart a service, scale a node, or ignore a false positive.
The workflow
What your AI does when an alert triggers.
The AI handles the heavy lifting of data retrieval and correlation.
Scan for active alerts
The AI looks through your Netdata Cloud spaces to find any currently active alarms.
list_space_alertsGet alarm details
It pulls the specific details for the alarms it found to understand the severity and the affected nodes.
get_alarmsFetch metric charts
The AI pulls the actual time-series data for the metrics involved in the alarm.
get_chart_dataSummarize findings
The AI looks at the chart trends and tells you exactly what happened on the node during the alert period.
get_chart_data
Try it
Copy these to start.
Use these prompts to kick off an investigation.
Starting points
Swap out the space names or node IDs with your actual Netdata identifiers.
Netdata Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_ahKhmIYu5E2SZYLAuopQ0XxxT8FYpWngo29KyMi2/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Netdata capabilities are ready to use.
{
"mcpServers": {
"netdata-mcp": {
"url": "https://edge.vinkius.com/vk_preview_ahKhmIYu5E2SZYLAuopQ0XxxT8FYpWngo29KyMi2/mcp"
}
}
}List all active alerts in my 'Production' space.
Check the details for the alarm on node 'web-server-01'.
Get the CPU and memory chart data for the last hour on node 'db-primary'.
Look at the recent alerts in 'Staging' and tell me which ones look like hardware issues.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Netdata once, then ask.
Log in to Netdata, connect the MCP, and your AI will immediately have access to your cloud spaces. All credentials stay encrypted.
Connect Netdata to your AIFAQ
How this task behaves.
- 01
Can the AI clear or acknowledge alarms in Netdata?
No. The AI can only read alert lists and metric data. You must manually acknowledge or clear alarms in the Netdata dashboard.
- 02
Can the AI change my Netdata configuration?
No. This MCP is read-only. It cannot change alarm thresholds, update node settings, or modify your Netdata Cloud setup.
- 03
Does the AI see all my Netdata data?
The AI only sees the data from the spaces you authorize when you connect the MCP.
- 04
How does the AI know which node is failing?
It uses the information provided by the alarm list to identify the specific node and then requests the relevant chart data for that node.
- 05
Can I use this with any AI client?
Yes, as long as your client supports the Model Context Protocol (MCP).
More questions about Netdata? See everything the Netdata Connector can do
Connect Netdata to Claude, Cursor, ChatGPT & more
