What problem does it solve? Teams building .NET services often lack consistent observability standards, resulting in ad-hoc metrics, missing alerts, and dashboards that cannot be compared across services. This Skill provides a single canonical standard for SLIs, alert thresholds, dashboard generation, and OpenTelemetry instrumentation so every service is monitored the same way. ## Core Features & Use Cases - Standard SLI Catalog: Predefined SLI tables with targets for HTTP APIs, background workers, and message consumers, covering latency percentiles, error rates, saturation, queue depth, and DLQ signals. - Alert Conventions: Severity levels (Critical, Warning, Info) with standard warning/critical thresholds and durations for error rate, latency, CPU, memory, queue, and worker metrics. - Grafana Dashboard Templates: Ready-to-adapt JSON dashboards (Service Health, API Performance, Resource Usage, Background Worker) with canonical datasource, env, and service template variables plus PromQL query examples. - OpenTelemetry Patterns: Constructor injection of ILogger, IDistributedTracing, and IMeterFactory, correct ActivityKind usage, environment attribution via deployment.environment, and a complete instrumentation code example. - Use Case: An Architect defining observability requirements for a new message-consumer service can pull the consumer SLI table, alert thresholds, and telemetry ownership rules directly, then hand the Developer the instrumentation pattern and dashboard generation rules to implement. ## Quick Start Ask the agent to define the SLIs, alert thresholds, and a Grafana dashboard for your .NET service using the observability skill.
