Datadog

The Datadog skill connects Claude to your Datadog environment for log search, metric queries, and monitor status. It covers the observability data you need for alert investigation and performance analysis.

Metadata

Field	Value
ID	datadog
Category	observability
Required env vars	DD_API_KEY, DD_APP_KEY
Optional env vars	DD_SITE (default: datadoghq.com)

Tools

Tool	Description
datadog_search_logs	Search logs using Datadog query syntax with time range filters
datadog_query_metrics	Query time-series metrics (e.g., avg:system.cpu.user{*})
datadog_list_monitors	List monitors, optionally filtered by name or tags

Setup

1. Go to Datadog > Organization Settings > API Keys and create (or copy) an API key.
2. Go to Organization Settings > Application Keys and create an application key.
3. Set the environment variables:

export DD_API_KEY="your-api-key"
export DD_APP_KEY="your-app-key"

Non-US regions

If your Datadog account is on a non-US site, set DD_SITE:

export DD_SITE="datadoghq.eu"        # EU
export DD_SITE="us3.datadoghq.com"   # US3
export DD_SITE="us5.datadoghq.com"   # US5
export DD_SITE="ap1.datadoghq.com"   # AP1

When DD_SITE is not set, the skill defaults to datadoghq.com (US1).

Key permissions

The application key inherits the permissions of the user who creates it. For SWEny, the user needs read access to logs, metrics, and monitors. No write access is required.

Workflow usage

Triage workflow:

• gather — Search logs for error patterns, query metrics for anomalies, and check monitor status to understand the scope of an alert.

Datadog is interchangeable with Sentry and BetterStack at the gather node. The executor uses whichever observability skills are configured. If you use multiple observability providers, Claude will query all of them for a more complete picture.