feat(monitoring): add Monitoring view mirroring concave-web

[devin-ai-integration]

Summary

Adds a Monitoring view to the Bubble Tea TUI that mirrors the Prometheus/Grafana surface now shipping in concave-web (Gradient-Linux/concave-web#1). Business logic stays in concave; the TUI only reads from upstream HTTP endpoints.

New package — internal/monitoring

• Prober with Probe(ctx) (reachability via /-/healthy + /api/v1/status/buildinfo and /api/health) and Query(ctx, expr) (PromQL instant, reduced to first scalar sample).
• Graceful degradation: empty URL ⇒ Configured:false; HTTP errors and decode failures propagate into the Sample.Error field so the view can render them inline.
• Covered by new unit tests in internal/monitoring/monitoring_test.go.

Config — cmd/concave-tui/config

• New [monitoring] TOML section with prometheus_url / grafana_url. Both default to empty; when both are empty the view renders a guidance card pointing at the config file and mentioning the Flow (:9090/:3000 bundled) and Forge suites.
• Covered by TestMonitoringURLsRoundTrip.

TUI — cmd/concave-tui/model

• New MonitoringModel between Fleet and Teams in the sidebar. Viewer-gated, refresh-on-r, auto-refresh every 15 s.
• Sidebar icon MN, label Monitoring; shortcut shifted from 8 to 9 for System on admin sessions.
• Two existing order-checking tests (TestRootSwitchesViewsWhenAuthenticated, TestAdminVisibleViewsIncludeSystemAndUsers, TestViewerVisibleViewsIncludeMonitoringScreens) updated to reflect the new view index — this is a direct consequence of the new sidebar entry, not a semantic change.
• New monitoring_test.go covers: guidance when unconfigured, reachability + PromQL rendering with a stubbed prober, unreachable-error rendering.

No new dependencies. go vet ./... and go test ./... pass.

Review & Testing Checklist for Human

Risk: yellow — adds a new sidebar entry (so view indices shift) and a new outbound HTTP path for probing Prometheus/Grafana, but both URLs are empty by default so the view is a no-op until the user opts in.

• Confirm TUI launches as before for users who have NOT edited ~/.config/concave-tui/config.toml — the existing [display] / [layout] content is untouched; a fresh [monitoring] block should be appended with empty URLs.
• With the Flow suite installed and running, set prometheus_url = "http://127.0.0.1:9090" and grafana_url = "http://127.0.0.1:3000", restart the TUI, switch to the Monitoring view (number 7 for viewers, or tab), and confirm the reachability badges flip to reachable and the PromQL rows populate.
• With only Prometheus configured (Grafana URL blank), confirm Grafana shows not configured rather than erroring.

Notes

• Monitoring URLs are intentionally NOT surfaced in the TUI settings modal — they are infrastructure-level concerns best edited in the TOML file alongside the other sysadmin-visible settings. Happy to wire them into SettingsModel as a follow-up if preferred.
• The key-number shortcut for System (and Users) shifts by one for admin sessions because the new view sits before them. The existing tab / shift+tab cycling continues to work as before.
• Related: feat(monitoring): add Prometheus and Grafana integration concave-web#1 (web monitoring surface).

Link to Devin session: https://app.devin.ai/sessions/5d19efa113054ca4953d9ed9309ce705
Requested by: @ElFariss

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring