A comprehensive bridge to Datadog's observability platform that exposes monitors, logs, metrics, traces, dashboards, and APM through MCP tools. You get grep-style log search, trace filtering with duration and error queries, and smart sampling modes to stay within token budgets. Supports both stdio for local Claude Desktop setups and HTTP transport for container or Kubernetes deployments. The monitor tools cover the full lifecycle from creation with schema validation to muting, unmuting, and tracking alert history with real state transitions filtered from renotify noise. Ships with Docker images and handles EU/US site selection. Useful when you need to query production observability data, correlate logs with traces, or manage alert configurations directly from an AI conversation without switching to the Datadog UI.
claude mcp add --transport stdio io.github.tantiope-datadog-mcp uvx datadog-mcp