# Console Observability

This document describes Console observability expectations: what telemetry matters, how to correlate UI actions with backend traces, and what to surface in air-gapped deployments.

## What to Measure (UI)

Recommended UI metrics include:

- **Time-to-verdict (TTFV):** from navigation to verdict banner rendered.
- **Time-to-evidence:** from clicking a fact/badge to evidence preview available.
- **Export latency and success rate:** evidence bundle generation time and failures.
- **Mute/exception usage:** how often operators suppress or escalate findings (counts, reversal rate).

## What to Log (Structured)

Console logs should be structured and tenant-scoped:

- `tenantId`, `actor`, `actionType`
- `artifactId` / image digest
- `findingId` / vulnerability identifiers (when relevant)
- `traceId` / correlation IDs that tie UI requests to backend traces

## Error Surfaces

Operators need actionable error messaging:

- Distinguish client validation errors from server failures.
- Provide a copyable correlation/trace ID for support.
- Avoid leaking stack traces or secrets into UI notifications.

## Offline / Sealed Mode Telemetry

In sealed mode, surface:

- snapshot identity and staleness budgets
- which data is stale vs fresh (policy pack version, VEX snapshot time, feed ages)

## References

- UI telemetry guidance: `docs/observability/ui-telemetry.md`
- Accessibility baseline: `docs/accessibility.md`