feat(ops): add data freshness dashboard

[Flickwire-Agent]

User Story

As the operator of the service, I want a data freshness dashboard so that I can quickly tell whether scans, caches, and external API ingestion are healthy.

Why This Matters

The app depends on scheduled scans across thousands of item-region combinations. If ingestion falls behind, users may make bad trading decisions. Operator-facing freshness visibility shortens incident detection and debugging.

Acceptance Criteria

• A dashboard or status page shows scan progress, oldest snapshot age, latest cache generation time, queue depth, failed jobs, and last successful worker activity.
• The dashboard distinguishes API health from market-data freshness.
• Freshness thresholds are visually classified as healthy, warning, or stale.
• The data is available through a documented API endpoint.
• The page avoids exposing secrets or sensitive infrastructure details.
• Existing /api/health and /api/worker/status behavior remains intact.

Useful Context

• /api/worker/status already exposes queue and 24-hour job stats.
• Market snapshot persistence can provide oldest/newest fetched timestamps.
• This is useful for both humans and future automated health checks.

Suggested Verification

• Build API and web packages.
• Manually compare dashboard values to worker status and database-derived timestamps.

[Flickwire-Agent]

Implemented in 74b8fee. Added documented, no-store GET /api/freshness plus the lazy-loaded /status operator view. It reports API dependency booleans, aggregate queue depth, current snapshot age bounds, last worker activity, and cache generation timestamps; each freshness signal is classified healthy, warning, or stale without exposing infrastructure details.\n\nVerified API and web builds, oxlint, oxfmt, the deployed endpoint, OpenAPI documentation, and the served /status route. Live status currently identifies snapshots as stale under the configured 6-hour policy while API dependencies, worker cadence, and both caches are healthy.