kurihiro0119/datadog-mcp-server

@kurihiro0119/datadog-mcp-server

MCP (Model Context Protocol) server for Datadog integration. Monitor metrics, logs, and APM data directly from your AI applications.

Features

• Alert Management: Retrieve and analyze active alerts
• Metrics Query: Query Datadog metrics with custom time ranges
• Log Search: Search and retrieve logs with filters
• Monitor Management: List and get details of monitors
• Alert Analysis: Automatic analysis of alerts with potential causes and suggested actions

Installation

npm install @kurihiro0119/datadog-mcp-server

Configuration

Set up your Datadog API credentials:

export DD_API_KEY="your-api-key"
export DD_APP_KEY="your-app-key"
export DD_SITE="datadoghq.com"  # or datadoghq.eu, us3.datadoghq.com, etc.

Usage

As a CLI tool

npx @kurihiro0119/datadog-mcp-server

In your MCP client configuration

{
  "mcpServers": {
    "datadog": {
      "command": "npx",
      "args": ["@kurihiro0119/datadog-mcp-server"],
      "env": {
        "DD_API_KEY": "your-api-key",
        "DD_APP_KEY": "your-app-key",
        "DD_SITE": "datadoghq.com"
      }
    }
  }
}

Available Tools

get_alerts

Retrieve active alerts from Datadog.

Parameters:

• limit (number): Number of alerts to retrieve (default: 10)
• priority (string): Filter by priority (P1-P5)
• tags (array): Filter by tags

get_alert_details

Get detailed information about a specific alert.

Parameters:

• alertId (string, required): The ID of the alert

get_metrics

Query Datadog metrics.

Parameters:

• query (string, required): Datadog metric query
• from (string): Start time (ISO 8601 or relative like "-1h")
• to (string): End time (ISO 8601 or relative)

get_logs

Search Datadog logs.

Parameters:

• query (string, required): Log search query
• from (string): Start time
• to (string): End time
• limit (number): Number of logs to retrieve (default: 100)

get_monitors

List Datadog monitors.

Parameters:

• tags (array): Filter by tags
• monitorTags (array): Filter by monitor tags
• name (string): Filter by monitor name

get_monitor_details

Get details of a specific monitor.

Parameters:

• monitorId (string, required): The ID of the monitor

analyze_alert

Analyze an alert and provide potential root causes and solutions.

Parameters:

• alertId (string, required): The ID of the alert to analyze
• includeMetrics (boolean): Include related metrics (default: true)
• includeLogs (boolean): Include related logs (default: true)

Example Usage

Once configured, you can use natural language to interact with Datadog:

• "Show me all P1 alerts"
• "Analyze the alert with ID 12345"
• "Get CPU metrics for the last hour"
• "Search logs for errors in the payment service"
• "List all monitors tagged with production"

Programmatic usage

// Query CPU metrics for the last hour
const result = await client.callTool('get_metrics', {
  query: 'avg:system.cpu.user{*}',
  from: 'now-1h',
  to: 'now'
});

Development

# Run in development mode
npm run dev

# Run tests
npm test

# Build for production
npm run build

License

MIT