Datadog MCP Server

Exposes Datadog's metric query API through MCP protocol, allowing Claude and other MCP clients to execute time-series queries against Datadog's metric backend. Translates MCP tool calls into authenticated Datadog API requests, handling query parameter serialization, time window specification, and metric aggregation options. Returns structured time-series data with timestamps and values for downstream analysis or visualization.

Enumerates all monitors configured in a Datadog account and retrieves their current status, alert state, and configuration details. Implements pagination to handle accounts with hundreds of monitors, supports filtering by monitor type (metric, log, APM, etc.), status, and tags. Returns structured monitor metadata including thresholds, notification channels, and last-triggered timestamps for decision-making.

Executes log queries against Datadog's log aggregation backend using Datadog's query language (DQL or legacy Lucene syntax). Supports full-text search, field-based filtering (service, environment, host, status code), time range specification, and result sorting. Returns paginated log entries with parsed fields, timestamps, and source metadata for investigation and analysis.

Queries Datadog APM (Application Performance Monitoring) to retrieve distributed traces and individual spans for a service. Supports filtering by service name, operation name, trace status (error/success), duration thresholds, and custom tags. Returns trace hierarchies with span timing, resource names, and error details for performance analysis and debugging.

Lists dashboards in a Datadog account and retrieves their full configuration, including widget definitions, metric queries, and layout information. Supports filtering by dashboard type (custom, service overview, etc.) and tags. Returns dashboard metadata and widget definitions in JSON format for analysis or programmatic dashboard generation.

Retrieves events from Datadog's event stream, supporting filtering by event type (monitor alert, deployment, custom event), source, tags, and time range. Returns event metadata including timestamp, title, text, and associated tags for timeline analysis and incident correlation.

Creates, updates, and lists downtime windows in Datadog, allowing suppression of alerts during maintenance or known issues. Supports recurring downtime schedules, scope filtering by monitor tags or specific monitors, and timezone-aware scheduling. Returns downtime configuration and status for audit and compliance tracking.

Submits custom metrics to Datadog via the metrics API, supporting gauge, counter, histogram, and distribution metric types. Handles metric naming, tagging, and timestamp specification. Enables programmatic metric generation from Claude-driven workflows for custom monitoring scenarios.

Retrieves service definitions from Datadog's service catalog, including service metadata, dependencies, owners, and tags. Supports filtering by team, environment, or custom tags. Returns service relationship information for understanding system architecture and ownership.

Creates and updates incidents in Datadog's incident management system, supporting severity levels, status transitions, and team assignment. Integrates with Datadog's incident workflow to enable programmatic incident creation from Claude-driven automation. Returns incident metadata and status for tracking.