Commit Graph

8 Commits

Author SHA1 Message Date
Anso 716daf77d0 feat(updates): auto-prune dangling images after updates (#1316)
* feat(updates): auto-prune dangling images after updates

Each update pulls a fresh image and recreates containers, leaving the
replaced image behind as a dangling layer that previously had to be
pruned by hand. A new "Prune dangling images after updates" toggle under
Settings > System > Docker hygiene reclaims these automatically.

The setting is on by default and opt-out. When enabled, a successful
stack update (manual or scheduled) and a Sencho self-update each remove
the dangling image layers they orphaned. Only untagged layers are
touched; tagged images, volumes, and data are never removed. The toggle
requires an admin account and is per node: each instance honors its own
value, so a remote node self-update applies that node's own preference.

A prune failure never affects the update result: on the stack path it is
caught and logged after the update has already succeeded, and on the
self-update path the helper-shell prune runs only after a clean recreate
and cannot change the exit code or the recorded update error.

* security(self-update): shell-quote label-derived values in helper command

Address review feedback on the prune-on-update change:

- The self-update helper command interpolated the compose service name and
  config-file paths (both read from Docker Compose labels) straight into a
  shell string. Shell-quote them via shQuote so a label carrying shell
  metacharacters stays inert data and cannot break the exit-code capture,
  error-file write, or prune guard.
- Correct the settings copy and docs: the prune is a standard dangling-image
  prune, so it reclaims every untagged layer on the node, not only the one the
  current update orphaned. Tagged images, volumes, and data remain untouched.
- Add tests: shell-metacharacter neutralization and prune-output suppression in
  the self-update command, and an atomic-update case asserting a prune failure
  does not trigger a rollback.

* fix(updates): omit the reclaim figure when the daemon reports zero bytes

End-to-end testing on a Docker daemon backed by the containerd image store
showed the post-update prune removing a dangling image while the prune API
returned SpaceReclaimed=0, so the stream printed "reclaimed 0.0 MB" even though
an image was removed. Show the reclaimed figure only when the daemon reports a
non-zero value; otherwise the line reads "=== Pruned dangling images ===". The
overlay2 store still reports real figures and shows them. Add a test covering
both branches.
2026-06-05 18:12:37 -04:00
Anso 0683aa9395 fix(settings): reflect role, tier, and node scope in settings panels (#1300)
* fix(settings): reflect role, tier, and node scope in settings panels

Three gate fixes so the Settings panels match what the backend enforces.

Non-admin roles saw editable fields and a Save button on System Limits,
Developer, and App Store, but writes require admin, so Save always failed.
These panels now render read-only for non-admins (controls disabled, Save
hidden) while still showing the values.

The Developer panel is node-scoped but read and wrote the controlling
instance regardless of the selected node, so a remote node's debug mode and
retention windows could not be changed from the UI. It now targets the
active node like System Limits.

The settings shell derived the Admiral entitlement locally; it now consumes
the backend-provided value the API authorizes against, and that value is
corrected to require an active paid tier so an expired Admiral license no
longer reports as Admiral.

* fix(settings): gate audit_retention_days writes behind Admiral

audit_retention_days configures the Admiral-only audit log (the audit-log
routes require Admiral, and the Developer settings UI only shows the field to
Admiral operators), but the settings POST/PATCH handlers only required an admin
role. A non-Admiral admin (for example Skipper, or an expired-Admiral admin
whose tier dropped to community) could still set it through the API.

Gate writes to that key with requireAdmiral on both the single-key POST and the
bulk PATCH paths, matching the audit-log routes and the UI. Other keys remain
writable by any admin.

* fix(settings): reject unknown keys on PATCH instead of silently stripping

The bulk settings PATCH validated the body with a Zod object schema that
strips unknown keys by default, so a request carrying a disallowed key (for
example an auth_* secret) returned 200 as a no-op instead of being rejected.
No secret was written, but it diverged from the single-key POST path, which
rejects disallowed keys, and could hide client drift.

PATCH now rejects any key outside the allowlist with a 400 before validation
or write, keeping the bulk path fail-closed and consistent with POST.
2026-06-04 01:49:23 -04:00
Anso aa3d99a594 fix(mesh): re-evaluate data plane every 10s and add opt-in auto-recreate (#1184)
* fix(mesh): re-evaluate data plane every 10s and add opt-in auto-recreate

MeshService.dataPlaneStatus was written exactly once at boot in
setupMeshNetwork() and never re-evaluated. After the operator removed
sencho_mesh at runtime (or it was recreated externally, or Sencho was
disconnected from it), /api/health and the dashboard banner kept
returning the stale boot-time discriminator until the next process
restart.

Adds a 10s revalidator that inspects the current Docker truth in one
network-inspect call and transitions dataPlaneStatus to reflect it.
Short-circuits in not_started / not_in_docker / subnet_invalid
(states that cannot change within this process) and in concurrent
ticks. Transitions are idempotent on stable state, so the timer can
tick indefinitely on a healthy mesh without log noise.

New 'not_found' reason value for the network-was-removed-at-runtime
case. Existing reasons (subnet_mismatch, subnet_overlap, attach_failed)
also surface from the revalidator when their underlying conditions
arise post-boot. transitionDataPlane keeps message and subnet fields
fresh across consecutive observations even when reason is unchanged,
so /api/health never reports stale numbers (e.g. two consecutive
subnet_mismatch observations against different external subnets).

Adds an opt-in mesh_auto_recreate global setting (default off). When
on, the revalidator additionally calls attemptInPlaceRecreate() after
surfacing not_found. The helper hard-prefers the boot-chosen subnet
(this.meshSubnet) and never iterates candidates, because changing the
subnet here would invalidate every existing extra_hosts override on
disk. A real conflict on the original subnet is reported as
subnet_overlap and preserved during the 60s recreate throttle window
so the operator-actionable reason is not flapped back to not_found
between attempts.

Self-attachment is checked via Name match (operator --hostname X
matches container Name /X) or full container-ID prefix for hex
HOSTNAMEs >= 12 chars (Docker default short ID). Non-hex HOSTNAMEs
cannot collide with container IDs at all so a Name miss is conclusive;
short hex HOSTNAMEs preserve the prior status as 'unknown' rather than
risking a false-positive prefix match.

Frontend surfaces:
- types/mesh.ts: 'not_found' added to MeshDataPlaneReason.
- MeshDataPlaneBanner: 'not_found' headline copy.
- Settings > System > Mesh data plane: TogglePill bound to
  mesh_auto_recreate, default off, helper text explains the tradeoff.

Backend coverage in backend/src/__tests__/mesh-data-plane-revalidate.test.ts
(25 cases): short-circuits, idempotent stable-state, recovery from
subnet_mismatch / subnet_overlap, transition to not_found / subnet_mismatch
/ attach_failed, transient-Docker anti-flap, re-entrancy guard, name
match path, ID-prefix path, short hex hostname ambiguity, non-hex
hostname certainty, transition message refresh on observation drift,
auto-recreate off (default), auto-recreate success with senchoIp
preservation, auto-recreate overlap classification with no subnet
drift, throttle window preserves classified reason, throttle release.
Lifecycle test covers timer wiring in start()/stop().

Existing mesh-setup-error-classification suite (27 cases) still green.

Resolves: F-4 in the v1.0 audit tracker.

* fix(mesh): address Codex review of PR #1184

Three findings from the independent review:

BLOCKER: attemptInPlaceRecreate() called recordSetupFailure() on
create / attach failures, which clears this.senchoIp. The next
revalidator tick's attachment check is guarded on senchoIp, so with
it null the check is skipped and the snapshot path can silently flip
the status back to ok against a network where Sencho is in fact not
attached. Also: a later successful recreate would call
ensureSelfAttached() with senchoIp null, which short-circuits, so
the network gets recreated without binding Sencho.

Replaced the recordSetupFailure() calls in attemptInPlaceRecreate
with a new recordRecreateFailure() that uses transitionDataPlane and
preserves senchoIp. Added two tests: create-fails-then-succeeds
(verifies senchoIp survives the failure and the later retry binds
Sencho correctly) and create-succeeds-attach-fails-then-next-tick
(verifies the snapshot path surfaces attach_failed on the next tick
instead of falsely reporting ok).

SHOULD-FIX 1: single-key POST /api/settings wrote String(value)
without re-validating against the per-key schema, so an allowlisted
enum-shaped key like mesh_auto_recreate could persist arbitrary
strings ('banana', 'true') that the bulk PATCH would later refuse.
Routed the single-key path through SettingsPatchSchema.safeParse so
both write paths validate identically. Added regression tests for
an invalid mesh_auto_recreate value, a valid mesh_auto_recreate
write, and an out-of-range numeric value.

SHOULD-FIX 2: the new Mesh data plane subsection lived inside a
section the registry exposes to non-admins, who would see the toggle
and only learn it was admin-only after the save 403'd. Gated the
subsection on `isAdmin` from useAuth so non-admins do not see the
control. The other system controls keep their existing visibility
pattern (read-only for non-admins).

71/71 backend tests green (revalidate + mesh-setup + settings-routes).
276/276 frontend tests green. tsc clean on backend + frontend.
2026-05-23 18:09:39 -04:00
Anso fcff8e9047 fix(monitor): collapse repeated host-metric alerts into per-window summary (F-11) (#1175)
* fix(monitor): collapse repeated host-metric alerts into per-window summary (F-11)

A host metric over threshold previously dispatched one notification every 5
minutes for the duration of the breach, producing 7+ identical messages
in 35 minutes and spamming Discord/Slack routes. Replace the hardcoded
5-minute cooldown for CPU/RAM/disk with a per-metric suppression window
(default 60 min, configurable via host_alert_suppression_mins). The first
breach fires immediately; subsequent cycles within the window are silently
counted; the next dispatch after the window elapses carries a summary
suffix listing how many cycles were suppressed and when the breach first
crossed threshold. Recovery clears the counter so re-breach fires fresh.

The pattern mirrors PolicyEnforcement.notifyTrivyMissingOnce: module-scope
Map, in-memory only, in-cycle dedup, with a test-reset helper. The
existing system_state row keeps post-restart re-fires bounded.

Janitor and per-stack alert rules are unchanged; they already have
adequate cadence and per-rule cooldown respectively.

* fix(ci): restore backend and frontend checks

* fix(e2e): remove create button timing race

* fix(e2e): harden create double-click test

* fix(monitor): clear persisted F-11 timestamp on recovery + clamp suppression window

Independent audit on the previous commit surfaced two issues.

1. clearHostMetricSuppression early-returned on missing in-memory state,
   leaving a stale system_state.last_host_*_alert_ts row alive after a
   process restart. Scenario: breach fires + persists timestamp, process
   restarts, metric recovers before another evaluate cycle re-seeds the
   in-memory Map, recovery cleanup early-returns. Next re-breach inside
   the original window hits the restart-survivability branch and is
   silently suppressed instead of firing fresh. Fix: read persisted
   state in clearHostMetricSuppression and reset to '0' independently
   of in-memory presence. The read-before-write also skips redundant
   writes when the row is already cleared.

2. host_alert_suppression_mins is validated by zod on the bulk PATCH
   path but the single-key POST /api/settings path accepts allowlisted
   keys without re-validation. A 999999999-minute value would silence
   host alerts for centuries. Add MAX_HOST_ALERT_SUPPRESSION_MIN = 1440
   mirroring the zod max, and clamp via Math.min in evaluateGlobalSettings.

Two new vitest cases (restart-then-recovery-then-rebreach; the 1440
clamp) confirmed failing before the fix, passing after. The existing
"metric drop" case updated to use a mock-backed persistence pattern
consistent with the new restart-scenario tests. 73/73 monitor-service
tests green; full backend suite 2507/2510 (same pre-existing Windows
EBUSY flake on filesystem-backup.test.ts as baseline).
2026-05-23 06:28:01 -04:00
Anso eead195529 feat(settings): dress the page to match the audit (#849)
* feat(settings): dress the page to match the audit (cyan rail, italic serif, two-column rows)

Brings the full-page Settings route into the Sencho voice. The page now
opens with a full-width PageMasthead (cyan rail, mono crumb, italic
serif title, contextual stat strip) above a sidebar and main-content
panel, each as a rounded-xl card inset on the dark background.

Sidebar drops the duplicate "Settings" header and the candy tier badges.
Group headers carry mono labels with visible/total counts; gated rows
get a neutral uppercase lock chip and dim. Active rows keep the cyan
2px rail.

Five new primitives (SettingsSection, SettingsField, SettingsCallout,
SettingsActions / SettingsPrimaryButton, TierLockChip) replace the
stacked label-input-help shadcn defaults and the per-section ad-hoc
chrome. AccountSection, AppearanceSection, LicenseSection, SystemSection,
NotificationsSection, DeveloperSection, AppStoreSection, AboutSection,
and SupportSection are migrated to the new layout. The list-driven
sections (Webhooks, Routing, Users, Labels, Security, CloudBackup,
ApiTokens, Registries, NodeManager, SSO) keep their list cards but get
the new chrome and primary CTAs.

Each section can publish contextual stats to the masthead via a small
context channel: 2FA state on Account, plan/trial/renews on License,
edited count on System, channel counts on Notifications, etc.

* refactor(settings): drop react-router-dom and align with DESIGN.md

The Settings page was the only surface using react-router-dom for sub-section
navigation. Every other primary view (Home, Fleet, Resources, App Store,
Schedules, etc.) drives view switching through a single activeView useState in
EditorLayout. This change removes the dependency end-to-end:

- App.tsx drops BrowserRouter
- EditorLayout adds 'settings' to the activeView union; SettingsPage renders
  inside the same flex-1 overflow-y-auto p-6 wrapper as siblings
- UserProfileDropdown receives an onOpenSettings callback instead of
  useNavigate. SettingsPage owns currentSection via props lifted to
  EditorLayout, so cross-component navigation (openLabelManager,
  onManageNodes, ConfigurationStatus rows) can route to a sub-section
- SettingsSidebar items become buttons (no more NavLink); SectionGate's
  redirect-on-invisible falls back through SettingsPage's safeSection memo
- e2e/nodes.spec.ts updates the Nodes selector from link to button role
- react-router-dom removed from package.json + package-lock.json

The visual treatment is brought into alignment with frontend/DESIGN.md,
which was rewritten this week to be the normative extract of the audit:

- PageMasthead: title text-3xl → text-[22px] Section rung italic; kicker
  11px → 10px Label rung; stat label tracking 0.22em → 0.18em; stat value
  font-medium for mono Stat-rung family discipline
- SettingsField helper: mono → sans Body rung 14/22; success tone now uses
  --success green (was incorrectly mapped to brand cyan)
- SettingsCallout: title tracking 0.18em; subtitle Body rung 14px; success
  tone now genuinely uses --success green; new brand tone for promotional
  callouts (Trial CTA, Admiral upgrade) that should read cyan
- SettingsActions: SettingsPrimaryButton renders mono uppercase tracked,
  size sm by default. DESIGN §9.10 requires "small mono uppercase, cyan-
  filled" for every Settings primary CTA
- TierLockChip: 9px → 10px Label rung floor
- SettingsSidebar: group header tracking 0.18em; ⌘K kbd 9px → 10px;
  aside gains text-card-foreground transition-colors per §10 canonical
  card class
- SettingsPage main panel: text-card-foreground transition-colors added;
  uses h-full overflow-auto p-6 to mirror FleetView's wrapper rhythm
- Field rows, section headers, action rows now consume var(--density-*)
  tokens with literal fallbacks so Settings respects the comfortable/
  compact toggle

* fix(e2e): update mfa openAccountSettings to match settings redesign

Settings now opens to the Account section by default when accessed from
the profile dropdown, and the Account section no longer renders an h2
heading element. Update the openAccountSettings helper to open the
correct section and assert on the Password h3 heading that SettingsSection
renders instead.

* test(e2e): fix MFA enrolment assertion after settings redesign

The 2FA enrolment badge was replaced with a kicker/field pattern.
Assert on the 'enrolled' text that the new design renders instead of
the removed Enabled badge.

* test(e2e): fix low-backup-codes warning assertions after settings redesign

Update two assertions in the 'low backup codes warning' test that
referenced UI text removed in the settings redesign:
- '1 backup code remaining' -> '1 remaining' (SettingsField body text)
- 'Regenerate now' button -> callout subtitle text, which uniquely
  identifies the zero-codes error card without hitting strict-mode
  from two identically-labelled Regenerate buttons on the page

* test(e2e): navigate to root before re-opening settings for mock refresh

The settings redesign uses a nested full-page route. Navigating to the
same URL a second time does not remount the component, so AccountSection
retains cached MFA state and the 0-codes branch never fetches. A
page.goto('/') ensures full unmount before the second openAccountSettings
call, so the refreshed mock is actually hit.

* test(e2e): scroll zero-codes callout into view before asserting visibility

The callout sits below the Disable 2FA section in the MFA settings page
and is scrolled out of the clipped content area on initial render.
scrollIntoViewIfNeeded() brings it into the visible viewport before the
toBeVisible assertion.

* test(e2e): scroll Radix ScrollArea viewport for zero-codes callout assertion

The settings page wraps content in a Radix ScrollArea whose Root has
overflow:hidden, so the browser's native scrollIntoView cannot scroll
the inner viewport. Wait for the callout to attach (confirms mock data
loaded), then programmatically set scrollTop on the Radix viewport
element before asserting visibility.

* test(e2e): use toBeAttached for zero-codes callout to avoid Radix clip issue

The callout renders below the Disable 2FA section, outside the visible
clip area of the Radix ScrollArea Root (overflow:hidden) on a standard
viewport. Playwright's visibility check uses the clip intersection, so
toBeVisible() fails even after programmatic scroll. toBeAttached()
confirms the component rendered the warning card for backupCodesRemaining:0
without depending on the element's scroll position.
2026-04-30 19:37:38 -04:00
Anso 9a1c043189 refactor(settings): replace modal with nested full-page route (#848)
* refactor(settings): replace modal with nested full-page route

Settings sections are now URL-addressable at /settings/:sectionId, rendered
nested inside EditorLayout alongside the stack sidebar. Browser back/forward
navigates between sections. Deep links (e.g. /settings/cloud-backup) load
the section directly on hard reload.

- Add react-router-dom v7; BrowserRouter wraps the full app tree
- New SettingsPage (scroll memory, Cmd+K palette), SettingsSidebar (NavLink
  active styling, back-arrow), SectionGate (visibility + tier lock card)
- Rename SectionId 'appstore' to 'app-store' so slug === SectionId
- Decouple SystemSection, DeveloperSection, AppStoreSection from modal-
  passed props; each fetches its own data on mount
- Replace onLabelsChanged prop chain with SENCHO_LABELS_CHANGED window event
- Drop onOpenSettings prop from UserProfileDropdown, HomeDashboard,
  ConfigurationStatus; each calls useNavigate directly
- Delete SettingsModal.tsx

* fix(settings): validate sectionId against registry before property write

Prevents prototype pollution (CodeQL js/remote-property-injection #243).
URL param sectionId is checked against SETTINGS_ITEMS before being used
as a property key on scrollPositionsRef.

* fix(settings): eliminate remote property injection via Map and registry-sourced key

Two-part fix for CodeQL js/remote-property-injection:

1. currentSection is now derived from SETTINGS_ITEMS.find().id (trusted
   registry data) instead of the raw sectionId URL param. The tainted
   string never flows into any property access.

2. scrollPositionsRef uses Map<SectionId, number> with .get()/.set()
   instead of a plain object. Map operations do not write to the prototype
   chain, removing the prototype pollution vector entirely.

* test(e2e): align settings selectors with full-page route

The settings refactor (4475afd) replaced the modal with a nested route.
The new sidebar renders sub-sections as NavLinks (role link, not button)
and adds a "Filter settings" button that collides with the loose
/settings/i regex used in mfa and nodes specs.

- Use exact 'Settings' match for the profile-dropdown menu row
- Switch the Nodes sub-section selector from button to link role
2026-04-30 12:57:02 -04:00
Anso 0bf061a745 feat(settings): group sections, add ⌘K search, scope breadcrumb (#680)
* feat(settings): group sections, add ⌘K search, scope breadcrumb

Restructures the Settings Hub sidebar into four labelled groups
(Identity, System, Alerts, Advanced), adds a ⌘K command palette for
section search, and surfaces the active scope (global vs node-scoped)
in the content breadcrumb.

- New `settings/registry.ts` centralises group/item metadata, tier gates,
  glyph assignments, visibility rules, and keyword hints consumed by both
  the sidebar and the command palette
- Cyan 2px left rail + gradient on the active sidebar item; mono-uppercase
  group headers; tier chips inline for locked items
- Scoped ⌘K handler via onKeyDownCapture on DialogContent so the hub no
  longer hijacks the global sidebar shortcut while open
- ScrollArea gains an opt-in `block` prop so the Nodes management table
  can overflow horizontally without Radix's default `display: table`
  wrapper clipping action buttons
- Docs reference updated with the grouped sidebar, scope breadcrumb, and
  ⌘K walkthrough plus refreshed screenshots

* refactor(settings): drop duplicate section headers, redesign system limits, always-visible tier chips

- Remove redundant section titles in every settings page; the dialog header now owns the title and description
- Rework System Limits into a compact row panel with inline-edit chips (warn state, focus ring) and an ON/OFF toggle pill
- Show tier chips on sidebar and command palette whether locked or unlocked, so Skipper/Admiral scope is always legible
- Keep right-aligned action buttons on pages that had a title+button header (Users, Labels, Nodes, API Tokens, Registries)

* fix(settings): seed NumberChip draft on edit instead of via effect

ESLint rule react-hooks/set-state-in-effect flagged the sync effect that
mirrored the external value into local draft state. Replace it with a
startEdit handler that seeds draft from value at click time, so the
button path always reads value directly and no cascading render is
triggered on prop change.

* fix(settings): restore heading role and clean sidebar accessible names

- Wrap the settings dialog title in an h2 so screen readers and E2E locators see a heading again after the in-section headers were removed
- Mark the sidebar glyph aria-hidden so the button's accessible name is just the item label (fixes anchored name matchers)
- Align the MFA E2E helper with the renamed Account section heading
2026-04-18 16:17:24 -04:00
Anso 7637091e84 feat(ui): glassmorphism redesign with settings decomposition (#274)
* feat(ui): add glassmorphism design tokens and utility classes

Introduce glass design system foundation: translucent oklch color
variables for both light and dark themes, glass/glass-border/glass-highlight
tokens, semantic status colors (success/warning/info), .glass and
.glass-strong utility classes with backdrop-filter, reduced shadow
values, and standardized spring animation presets in lib/motion.ts.

* feat(ui): apply glass treatment to core components

Update card, dialog, input, button, popover, sheet, tooltip,
dropdown-menu, context-menu, select, alert-dialog, and tabs
components with glassmorphism styling: translucent backgrounds via
new CSS variables, backdrop-blur layers, glass-border luminous
edges, and glass-highlight hover states.

* refactor(settings): decompose Settings Modal into section components

Extract 10 inline sections from the 1,987-line SettingsModal into
dedicated files under components/settings/. Introduce section
registry pattern replacing 14 conditional blocks. Add shared types,
sidebar navigation grouping with separators, glass treatment on
sidebar and nav buttons, and responsive modal height. SettingsModal
shell shrinks to ~380 lines.

* refactor(ui): unify all tabs to animate-ui TabsHighlight with glass styling

Migrate 4 tab instances (EditorLayout, FleetView, ResourcesView,
NotificationsSection) from inconsistent patterns (manual layoutId,
underline border-b-2, default fade) to the shared TabsHighlight
primitive with glass-highlight indicator and springs.snappy
transition. Standardize EditorLayout nav highlight spring config,
apply glass-highlight to sidebar stack list hover/active states,
and update mobile nav styling.

* refactor(ui): migrate hardcoded colors to semantic CSS variables

Replace hardcoded Tailwind color classes across ~19 component files
with semantic CSS variable classes: emerald/green to success,
orange/amber to warning, blue to info. Preserves brand/decorative
colors (Crown amber, Admiral blue). Enables consistent theming of
status indicators across the entire application.

* refactor(ui): Linear dark precision aesthetic — solid surfaces, depth cues, text hierarchy

Replace glassmorphism with Linear.app-inspired design: solid surface tokens
(card #111111, sidebar #0d0d0d, root #0a0a0a), backdrop-blur restricted to
floating overlays only (blur(10px) saturate(1.15)), desaturated teal accent,
font-weight 500 everywhere, monochrome chart palette, and three depth cues:
root ambient glow, luminous card top-edge, steep text brightness ramp.

* refactor(ui): precision polish — fix muddy dark, snowblind light, add design anchors

- Replace 34 hardcoded rgba values with theme-aware stat-* CSS tokens
- Fix light theme: solid white cards, off-white background, readable text
- Add card-border tokens with sharper directional lighting (top edge 2x)
- Add chart-grid/chart-tick tokens for theme-aware axis rendering
- Upgrade body glow: teal-tinted (dark), warm amber (light)
- Terminal-inspired sidebar: monospaced UP/DN status codes, Geist Mono
- Add tabular-nums to stat values to prevent layout jitter
- Light mode cards get shadow-sm for depth against off-white background

* refactor(ui): Linear materiality pass — ghosted nav, translucent sidebar, font unity

- De-escalate Delete button from solid destructive to ghost with hover fill
- Make sidebar translucent (80% opacity + backdrop-blur) so body glow bleeds through
- Bump dark nav accent to 0.07 for ghosted backlit selection
- Unify all terminal/editor fonts to Geist Mono (was JetBrains/Consolas mix)
- Add Monaco editor fontFamily for YAML/env editing consistency
- Add threshold-based color to Host RAM and Host Disk stat values (warn/crit)

* refactor(ui): material simulation — inherent depth, layer separation, recessed terminal

- Bump dark background 0.065→0.08, card surfaces 0.10→0.12 for 4% layer separation
- Add card-bevel token (inset top shimmer) for permanent structural depth
- Add button-inner-glow token for physical key feel on outline buttons
- Recess terminal with inset shadow and dimmed label
- Reduce action icon strokeWidth to 1.5 for refined industrial feel
- Add teal LED backlight bar on active nav item via blur pseudo-element

* fix(ui): parse usagePercent string to number for getValueColor

usagePercent is typed as string in SystemStats but getValueColor
expects number, causing TS2345 in CI builds.
2026-03-30 12:48:42 -04:00