Datadog MCP Server

A Model Context Protocol (MCP) server that enables AI assistants to search and analyze Datadog logs directly. This integration allows AI assistants to query your Datadog logs using natural language, making troubleshooting and log analysis more efficient.

Features

Log Searching: Search Datadog logs using flexible query parameters
Time-Based Filtering: Filter logs by specific time ranges
Pagination Support: Navigate through large log sets with cursor-based pagination
Coming Soon: Metrics retrieval functionality

Prerequisites

A Datadog account with API access
Datadog API and Application keys(See docs for detail)

Setup

Add the MCP server configuration to your AI assistant's MCP settings file:

Note: This MCP server only supports stdio transport. SSE transport is not supported.

"datadog-mcp-server": {
  "command": "npx",
  "args": ["-y", "@i524/datadog-mcp-server"],
  "env": {
    "DD_API_KEY": "<your-datadog-api-key>",
    "DD_APP_KEY": "<your-datadog-app-key>"
  }
}

Usage

Available Tools

search-logs

Search logs from Datadog with flexible filtering options.

Parameters:

filterQuery (required): A query string to filter logs. See Datadog Log Search Syntax for details.
filterFrom (optional): The minimum timestamp for requested logs (ISO 8601 format).
filterTo (optional): The maximum timestamp for requested logs (ISO 8601 format).
cursor (optional): The cursor to use for pagination.

Example Usage:

// Search for error logs from a specific service in production
{
  "filterQuery": "service:awesome-service status:error",
  "filterFrom": "2025-01-01T00:00:00Z",
  "filterTo": "2025-01-02T00:00:00Z"
}

License

This project is licensed under the MIT License - see the LICENSE file for details.