administration

OpenTelemetry integration exports Claude Code metrics—sessions, tokens, costs—and events like prompts and tool results to monitoring backends. Configure an exporter (OTLP, Prometheus, or console), set export intervals, and segment metrics by user, account, session, or model to track adoption and spending patterns across your deployment.

Track session costs by model and user, token usage, lines of code generated, tool adoption rates, and prompt frequency. Use these metrics with ROI templates to calculate productivity gains, quality improvements, and organizational adoption rates, enabling data-driven decisions on model selection and budget allocation.

Yes. Configure OpenTelemetry to export metrics via the Prometheus exporter, then visualize session costs, token usage, and adoption trends in Grafana dashboards. This setup identifies high-usage areas and helps optimize model selection and spending.

Enable OpenTelemetry telemetry, export cost metrics segmented by user and model, and configure alerts in your monitoring system—Prometheus or similar—based on session cost thresholds. This flags overspending before it escalates.

This Skill targets telemetry-enabled Claude Code deployments across multiple users, organizations, sessions, and models. It supports any deployment where you need visibility into usage patterns, cost control, compliance tracking, and ROI measurement over time.

No. OpenTelemetry supports multiple exporters: console output for quick testing, OTLP for existing observability platforms, or Prometheus for self-hosted stacks. Start with console export to validate metrics, then integrate with your preferred backend.