Datadog MCP Server
Datadog MCP servers enable AI models to interact with Datadog observability: metrics, logs, traces, monitors, dashboards, incidents, and infrastructure insights.
Overview
The Datadog MCP Server bridges AI agents with Datadog by providing structured access to observability data and controls. It enables natural-language workflows over metrics, logs, traces, dashboards, monitors, incidents, and infrastructure contexts.
Implementations:
Official preview by Datadog and community servers in Python, Node.js, and Docker.
Key Features
Metrics & Logs
Query timeseries metrics and search logs with filtering and pagination
Monitors & Alerts
List and inspect monitor states for alerting and SLO overview
Dashboards & Incidents
Discover dashboards and fetch incidents for operational context
APM & Traces
Access trace data for latency, dependencies, and service analysis
Available Tools
Quick Reference
| Tool | Purpose | Category |
|---|---|---|
get_metrics | Query timeseries metrics | Read |
search_logs | Search logs with filters | Read |
get_monitors | Retrieve monitor states | Monitoring |
list_dashboards | List dashboard definitions | Discovery |
get_incidents | List incidents | Incident |
Detailed Usage
get_metrics▶
Query Datadog metrics with flexible time ranges.
use_mcp_tool({
server_name: "datadog",
tool_name: "get_metrics",
arguments: {
query: "avg:system.cpu.user{*}",
minutes_back: 30
}
});
search_logs▶
Search logs with query, time window, pagination, and sorting.
use_mcp_tool({
server_name: "datadog",
tool_name: "search_logs",
arguments: {
query: "service:api-gateway AND status:error",
minutes_back: 30,
limit: 50,
sort: "-timestamp"
}
});
get_monitors▶
Retrieve monitor states with optional filters.
use_mcp_tool({
server_name: "datadog",
tool_name: "get_monitors",
arguments: {
groupStates: ["alert", "warn"]
}
});
list_dashboards▶
List dashboard definitions for discovery.
use_mcp_tool({
server_name: "datadog",
tool_name: "list_dashboards",
arguments: {}
});
get_incidents▶
List incidents with optional filtering and pagination.
use_mcp_tool({
server_name: "datadog",
tool_name: "get_incidents",
arguments: {
query: "state:active",
pageSize: 10
}
});
Installation
{
"mcpServers": {
"datadog": {
"command": "npx",
"args": [
"datadog-mcp-server",
"--apiKey", "your_api_key",
"--appKey", "your_app_key",
"--site", "datadoghq.com"
]
}
}
}
Regional Sites:
Use your Datadog site, e.g. datadoghq.eu, us3.datadoghq.com, us5.datadoghq.com, ap1.datadoghq.com.
Sources
Related Articles & Guides
Browserbase MCP Server
Browserbase MCP server enables AI models to control cloud browsers with Stagehand AI, providing automated page navigation, data extraction, element observation, and web actions.
Browserless MCP Server
Browserless MCP server provides AI-controlled cloud and self-hosted headless browsers with smart scraping, crawling, PDF generation, and anti-bot protection for Puppeteer and Playwright.
Docker MCP Server
Manage Docker containers, images, networks, and volumes through AI assistants using natural language with the Docker MCP Server.