Datadog

This plugin is currently in Preview.

Datadog Claude Code Plugin

Query your Datadog data directly from Claude Code using natural language. Ask about logs, metrics, traces, dashboards, monitors, and more.

What you need

• A Datadog account
• Claude Code (v2.1.30+)

Getting started

If you already have the Datadog MCP server registered separately (e.g., in .claude/mcp.json), disable or remove it first to avoid conflicts. Run /mcp in Claude Code, then restart Claude Code.

Start Claude Code, install the plugin from the official marketplace:

/plugin install datadog@claude-plugins-official

Auto-updates: Enable auto-update so Claude Code notifies when an update is available. Run /plugin, select the Marketplaces tab, select claude-plugins-official, then select Enable auto-update.

Before you can start querying your Datadog data, you’ll need to connect the plugin to Datadog using your account. The setup process will guide you in selecting the correct Datadog MCP domain. After setup, follow the instructions shown in Claude Code.

You can manually trigger setup by running the /ddsetup command.

Using the plugin

Once connected, just ask the agent anything about your Datadog data:

Show me error logs for the "checkout" service from the last hour

What monitors are currently alerting?

Find traces for service "api-gateway" with latency > 500ms

List my dashboards

Can't connect?

Never connected before? Run the /ddsetup command. It will help you provide the correct Datadog MCP domain and set up the MCP server.

Was working before but stopped? Run the /ddconfig command. It will check your site, authentication status, and network access to help diagnose the issue.

Changing settings

The plugin provides a few commands you can run in the agent to manage configuration:

• /ddconfig — change your Datadog site or switch organizations
• /ddtoolsets — enable or disable groups of tools
• /ddviz — enable, disable, or check the status of the visualization panel

Advanced usage

Key authentication

Instead of OAuth, you can authenticate using a Datadog API key and application key. Set all three environment variables before starting Claude Code:

DD_MCP_DOMAIN=your-mcp-domain \
DD_API_KEY=your-api-key \
DD_APPLICATION_KEY=your-application-key \
claude

The DD_MCP_DOMAIN value must be the MCP domain (e.g. mcp.datadoghq.com, mcp.us3.datadoghq.com, mcp.datadoghq.eu), not a URL — do not include https://. When using key authentication, /ddsetup is not required — the plugin connects directly.

Environment variable overrides

The plugin uses environment variables with default values in its registration file. You can override these defaults by setting the environment variables directly:

• DD_MCP_DOMAIN — overrides the Datadog MCP domain. If set, the plugin uses this value regardless of what /ddsetup or /ddconfig configured. Useful for non-standard environments or key authentication.
• DD_MCP_TOOLSETS — overrides the enabled toolsets (comma-separated). If set, the plugin uses this value regardless of what /ddtoolsets configured.

When environment variables are set, /ddsetup, /ddconfig, and /ddtoolsets still edit the default values in the registration file, but those defaults won't take effect until the environment variables are removed.

Good to know

• By default, authentication is handled via OAuth in your browser. Key authentication is also supported.
• No Datadog credentials are sent to the AI model provider.
• On macOS, visualization tool results can render in a native side panel. Run /ddviz to enable, disable, or check its status. Disabling applies to every Claude Code install sharing the default data directory (a dev build with DDVIZ_DATA_DIR set is controlled separately).
• When the panel can't render, the plugin sends lightweight telemetry (a failure-reason code only, never your queries or results) so we can spot broken setups. Opt out with DO_NOT_TRACK or DISABLE_TELEMETRY.

Support

• Datadog MCP Server Documentation

Legal

See the LICENSE and NOTICE files included with this plugin.

For details on how Datadog handles your data, see the Datadog Privacy Policy.