datadog-auto-detector

Datadog auto-detector recognizes Datadog URLs, service names, and natural-language observability queries in your input, then automatically retrieves and summarizes logs, metrics, traces, and related resources. You can ask "Check error rate for service-name" and get condensed data instantly without building queries manually.

Yes. The skill classifies your intent—whether you need full context, a specific query result, quick status, or comparative analysis—then fetches precisely matched data from logs, metrics, traces, monitors, and incidents. This accelerates debugging by eliminating context switching and query construction overhead.

The skill detects full Datadog URLs, natural-language observability queries like "error rate of service over the last hour," and service-name references in monitoring contexts. It then classifies intent and retrieves summarized observability data from logs, metrics, traces, dashboards, monitors, and incidents.

Yes. It integrates through the mcp-datadog dependency and operates via a three-tier architecture (skill → command → subagent) to isolate context. The skill fits into observability workflows by automating data retrieval so you focus on analysis and problem-solving rather than navigation.

Auto-detector eliminates manual query building and dashboard navigation by recognizing observability intent from natural language or URLs. It fetches condensed, AI-optimized summaries of logs, metrics, and traces automatically, reducing debugging friction compared to manual Datadog UI interactions.

Yes. The skill classifies comparative analysis as an intent type and orchestrates multi-resource retrieval to deliver summarized context. You can ask it to compare error rates, latency, or other metrics across services or time windows without manual data consolidation.