Reflect tracks token usage, costs, and session activity for monitoring and debugging.

Token usage is tracked per session and aggregated.

reflect usage
reflect doctor    # Full health report

Usage shows:

• Input/output tokens per model
• Total cost estimate
• Session count

Reflect can send anonymous usage telemetry.

// ~/.reflect/settings.json
{
  "telemetry": {
    "enabled": false
  }
}

Or:

reflect --telemetry    # Enable
haw --no-telemetry  # Disable

Export usage to your own OTEL collector:

{
  "telemetry": {
    "otel_enabled": true,
    "otel_endpoint": "https://collector.example.com:4318"
  }
}

Set OTEL_EXPORTER_OTLP_HEADERS for authentication.

Per-session metrics include:

• Turn count
• Tool call count
• Token totals
• Duration

Accessible via /ecosystem in TUI.

Reflect Cloud provides managed usage tracking:

• Organization-level metrics
• Budget alerts
• Entitlement management

See graycode-cloud documentation for details.

© 2026 GrayCode AI. All rights reserved.