Verified

Datadog MCP Server

Datadog API coverage with 163 tools for full SLO CRUD and fleet automation.

Data last scanned today · Reviewed today

53 12 dangerous tools106 tools★ 2TypeScriptMIT

Overview

This server provides a deep integration with Datadog, offering 163 distinct tools to manage monitoring, SLO lifecycles, and fleet deployments. It distinguishes itself by aggregating multiple sequential API calls into single structured responses, allowing users to move beyond simple data retrieval to complex diagnostic workflows. Built on the @us-all MCP standard, the server features tools for full SLO CRUD, status page management, and automated deployments. It supports efficient operation through token-saving field projection and granular tool toggles, with both stdio and Streamable HTTP transport modes available.

Our verdict

With a security score of 38/100, this server is currently high-risk and should be used with caution in sensitive environments. The low score reflects that the project is community-maintained rather than vendor-official and has not undergone an independent editorial review of its authentication or read-only modes. It is best suited for developers or operators working in non-production environments who require its extensive fleet automation and CRUD capabilities.

Provides 163 tools including full SLO CRUD functionality.
Aggregates complex API sequences into simplified structured responses.
Includes fleet automation tools for deployments and schedules.
Efficient token usage via search meta-tools and field projection.

Not maintained or officially supported by Datadog.
Security implementation remains unverified and not editorially reviewed.
Relies on a small contributor base increasing bus-factor risk.

Setup

Claude Desktop

{
  "mcpServers": {
    "datadog-mcp-server": {
      "args": [
        "-y",
        "@us-all/datadog-mcp"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

No read-only mode — running this grants full read/write access.
No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

Tool	Description	Risk
query-metrics	Query metrics data from Datadog.	low
get-metrics	Retrieve metric information.	low
get-metric-metadata	Retrieve metadata for a specific metric.	low
list-active-metrics	List currently active metrics.	low
list-metric-tags	List tags associated with a metric.	low
get-monitors	List all monitors.	low
get-monitor	Get details of a specific monitor.	low
create-monitor	Create a new monitor.	medium
update-monitor	Update an existing monitor.	medium
delete-monitor	Delete a monitor.	high

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›What authentication is required to run this MCP server?

You must provide both a Datadog API key and a Datadog Application key via the DD_API_KEY and DD_APP_KEY environment variables.

›Are write operations enabled by default?

No, write operations such as creating, updating, or deleting SLOs are disabled by default. You must set the DD_ALLOW_WRITE environment variable to true to enable these mutations.

›How can I reduce token usage given the large number of tools?

You can use the DD_TOOLS allowlist to load only specific categories of tools or the DD_DISABLE denylist to exclude unwanted ones. Additionally, the server includes a search-tools meta-tool to help manage context efficiently.

›Does this server support remote connections?

Yes, while it defaults to stdio transport, you can set MCP_TRANSPORT to http to use Streamable HTTP. When using HTTP mode, you must also provide an MCP_HTTP_TOKEN for Bearer authentication.

›Can I use this server with Datadog sites other than the default?

Yes, you can configure the specific Datadog site using the DD_SITE environment variable, which defaults to us5.datadoghq.com if not specified.