azure-resource-health-diagnose

Identify the resource by name, then assess its provisioning state, availability, and utilization. Query Log Analytics or Application Insights with KQL for errors and performance anomalies, classify findings by severity, and build a phased remediation plan with Azure CLI commands.

Query the requests table filtered to failed responses over the last 24 hours, summarizing failure counts by result code per hour. Correlate spikes with recent deployments or configuration changes to isolate the root cause before applying fixes.

The workflow uses the Azure MCP server tools (azmcp-*) when available, falling back to the Azure CLI. The target resource must be deployed and emitting telemetry through Log Analytics or Application Insights for log-based diagnosis to work.

No meaningful log analysis is possible without telemetry. The recommended action is to enable diagnostic settings on the resource, wait for data to accumulate, then rerun the assessment; only a generic health evaluation is possible otherwise.

A name alone can match multiple resources across subscriptions or resource groups. The workflow searches all subscriptions and, when multiple matches exist, asks you to specify the subscription or resource group to disambiguate.

No. The workflow presents a health assessment summary and asks for explicit confirmation before generating the detailed remediation report. Corrective actions are only executed after the user approves the proposed plan.