Multi-agent review of the feedback batch (7 confirmed):
- [P2] Low-battery sweep excluded a genuine 0% reading (`$pct > 0`) — a dead
battery never alerted. Now `< 20` (missing reading still defaults to 100).
- [P2] SMTP encryption rule `required|in:,tls,ssl` rejected the "None" option it
offered → unencrypted relays uncofigurable. Now `in:,tls,ssl`.
- [P2] Avatar preview called temporaryUrl() on any upload → 500 for HEIC/
non-previewable files. Guarded with isPreviewable() + real-time
updatedAvatar() validation for a friendly error.
- [P3] NotificationService::send() now catches transport errors so one SMTP
failure can't abort a whole sweep / automation run.
- [P3] Mailer config is now purged (Mail::purge) so long-running Horizon workers
pick up SMTP setting changes instead of reusing a cached mailer.
- [P3] Alert dedup TTL 1 day → 1 month, so a persistent outage mails once, not
daily.
- User: themed .checkbox (accent, dark-surface, custom check) replaces the
default browser checkbox in SMTP/automation/device modals.
+2 tests (0% battery alerts, no-encryption saves). Suite 84 green, 12/12 clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Issue 4: an SMTP add-on for e-mail notifications.
- SmtpSetup modal: server (host/port/encryption/user/pass), from/to, and
individual toggles for the events you want mailed — device offline, low
battery, automation messages. "Send test e-mail" button. Config stored
encrypted on the addon row; a runtime mailer is built from it (no .env edits).
- NotificationService: notify(event, …) sends only if SMTP is set up AND that
event is enabled; send() for the test.
- Triggers: automation "notify" actions e-mail (when enabled);
notifications:sweep (every 5 min) e-mails on NEW device-offline / low-battery
with per-condition dedup + battery hysteresis, so it's one mail per event.
6 tests (gating, per-event toggle, config persist + password-keep, sweep dedup).
Suite 82 green, 12/12 tabs clean; card + modal browser-verified.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Issue 2: Ring sat on "verbinde…" forever because nothing ever connected — the
ring-mqtt bridge (opt-in container) wasn't running, and HomeOS was collecting
Ring credentials it never used. Per the handoff the BRIDGE owns the Ring login
(email/pw/2FA) in its own web UI; HomeOS must not handle Ring credentials.
- Replaced the credential modal with a RingSetup guide: start the bridge
(`docker compose --profile addons up -d ring-mqtt`), open its web UI (linked
at the auto-detected host:port), log in there; devices then appear.
- Status is now honest — it reflects the bridge's REAL MQTT presence
(IngestRingMessage.markStatus), showing "Nicht verbunden" with next-steps
until the bridge actually reports, instead of a stuck "connecting".
- Removed the RingConnect cred modal + fields; reset any stuck status.
Suite green, 12/12 tabs clean; setup modal browser-verified.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Issue 1 — presence was half-built:
- Away detection: the debounce was 8 min (you disconnect, still "home"). Now
3 min and configurable (homeos.presence.away_debounce_minutes /
PRESENCE_AWAY_DEBOUNCE).
- Person management: ManagePerson modal handles BOTH add and edit (name, UniFi
client, avatar). Delete from the person card (confirm). Replaces AddPerson.
- Avatar upload: image stored on the public disk (storage:link), shown on cards
+ the who's-home strip; initials fallback (green when home).
- Who's home: a summary strip ("N Zuhause · M abwesend" with avatars) + clear
per-person home/away, so you can see at a glance who is in.
5 tests (debounce, create+avatar, edit, delete+avatar removal). Suite 78 green,
12/12 tabs clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Issue 5: rooms couldn't be created or deleted.
- "Raum anlegen" button + CreateRoom modal (name + icon picker) on the rooms
index.
- Delete on the room page behind a confirm that states how many devices are in
it; on delete the devices are KEPT but moved to "no room" (room_id null).
3 tests. Suite green, 12/12 tabs clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Issue 3: the New-Automation modal was taller than the screen and the whole
dialog scrolled. New <x-modal> component: header + footer are fixed (shrink-0),
only the body scrolls, capped to max-h-85dvh so the page never scrolls. Refactored
create-automation (now a responsive 2-column grid, widened to 3xl), add-device,
ring-connect and add-person onto it — the standard for every modal going forward.
Verified: modal fits a 720px viewport with no page scroll; create flow still works.
12/12 tabs clean, suite green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Issue 6 (2-3s toggle lag): the command service now writes the expected state
immediately for ALL devices (not just demo). The card flips within the
Livewire round-trip; the real MQTT echo / post-command HTTP poll reconciles it
(monotonic — newer observed_at wins). Toggling now feels instant.
- Issue 7 (wrong icon): Device::displayIcon() with a per-device override
(config.icon) + a picker on the device page (lamp/led/plug/window/door/
doorbell/sensor/…); cards use it. New icons added.
- Issue 8 (dead "Update prüfen"): real Shelly.CheckForUpdate over HTTP now
reports available version / up-to-date / unsupported instead of a demo string.
- Issue 9 (no MQTT option): device page can switch a Shelly between Local (HTTP,
probed) and MQTT transport; driverFor already routes by protocol.
Suite 71 green, 12/12 tabs clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
You were right that HA uses the Shelly LOCAL API, not MQTT. Adds that path
(user chose "both") — control + status over http://<ip>/rpc, no MQTT setup on
the device:
- ShellyRpc (POST /rpc), ShellyHttpDriver (Switch/Light.Set, Reboot).
- ShellyStatusApplier: one shared apply path (normalize + input roles +
monotonic upsert + broadcast) reused by BOTH the MQTT ingest and the HTTP
poll, so transports can't drift. IngestShellyMessage refactored onto it.
- ShellyLocalOnboarder: probe an IP → GetDeviceInfo/GetStatus → create an
http-protocol device with its entities (reuses an MQTT-onboarded row by id,
no duplicate). AssignDevice uses it when a discovered Shelly is reachable;
falls back to MQTT-style if not.
- Manual "Gerät hinzufügen" modal (add by IP). shelly:poll scheduled every 10s
+ a re-poll after each command (PollShellyDevice) for near-live status.
driverFor picks http vs mqtt by protocol.
- Normalizer now drops housekeeping components (sys/wifi/cloud/mqtt/ws/…) so
GetStatus doesn't create junk entities.
6 ShellyHttpTest cases (Http::fake). Suite 68 green, 12/12 clean.
LIVE-VERIFIED against the real Shelly 1 Mini Gen3 at 10.10.30.78: onboarded
over local API (protocol http), entities switch:0 + input:0, kept online by the
10s poll — no MQTT configured on the device.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Addresses the device-management feedback:
- The MQTT server address is no longer hardcoded (was 10.10.90.110). It's
auto-detected from the address you open HomeOS at (App\Support\HostAddress —
request host + port), so it follows a DHCP/changed IP. MQTT_DEVICE_HOST is now
an optional override only. Settings + device page use it.
- Discovery dedup: a finding whose prefix already belongs to a device (assigned
OR auto-onboarded) no longer shows under "Neue Geräte" — fixes the device
appearing both in discovery and in Geräte.
- Delete device: confirm-guarded button on the device page; deleting frees the
discovery finding so the device can be re-added.
- "Neu scannen" button on the network page → publishes homeos/sidecar/rescan;
the sidecar (now a subscriber) re-queries mDNS. ACL grants it read on that
topic. paho-mqtt pinned to the v2 callback API.
+4 tests (host override/fallback, dedup, delete frees finding). Suite 62 green,
12/12 clean. Live-verified: rescan reaches sidecar; duplicate Shelly gone.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Multi-agent review (5 confirmed) + Codex found real issues in the rework:
- [P1 security] Shared `shelly` ACL had readwrite +/rpc, so a compromised
device could inject Switch.Set into ANY other device's <prefix>/rpc. Now
read-only on +/rpc (receive own commands) + write only homeos/rpc (reply).
Broker-verified: a shelly publish to victim/rpc is denied; laravel's is not.
- [P2 security] `+` wildcard reaches reserved homeos/ring namespaces → bogus
device. Ingest now rejects RESERVED_PREFIXES (homeos/ring/$SYS).
- [P2 security] Unbounded auto-onboarding = DB-exhaustion DoS. Added a device
cap (homeos.mqtt.max_devices, default 250).
- [P2 correctness] Every Shelly `input` became a phantom window contact (wall
switches shown as windows, possibly inverted). `input` is now a generic
sensor; the user PROMOTES specific inputs to window/door contacts on the
device page (invert-aware), stored in config->input_roles and applied in the
ingest — this is the "assign contacts" flow the user asked for.
- [P3 ux] Motion pill read `active`; producer writes `on`. Now reads both.
- [P1 migrations] Dedup computed survivor keys once; 3+ duplicates could
collide on unique(device_id,key). Re-query per duplicate (ring + mqtt_prefix).
+9 tests (reserved prefix, cap, input generic/promoted/inverted, demo echo).
Live-verified: input published → onboarded as input → assigned window contact
via UI → appears on Fenster page, persists across messages. Suite 57 green,
12/12 tabs clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Addresses "I arrange tiles but nothing changes":
- Demo devices have no hardware to echo a new state, so toggling looked dead.
DeviceCommandService now simulates that echo for demo devices (writes the new
state + broadcasts DeviceStateChanged) — the mock home is fully interactive
(handoff §13.2). Real devices are untouched; their own status message applies.
- Panel tiles show the entity name (device · room as sub) and drive the shared
TogglesEntities toggle; wire:loading guards double-taps.
- Drag-reorder now confirms with a "Layout saved" cue so it doesn't feel inert.
+1 test (demo toggle simulates echo; real device does not). Browser-verified:
tapping an "Aus" tile flips it to "An". Suite 52 green, 12/12 clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Addresses "the list on every page isn't good": adopt the approved mockup's
card design (icon well, name, sub, live state chips, footer toggle).
- New <x-device-card>: per-device card with online dot, cloud badge, entity
chips and an inline switch/light toggle (active cards highlighted accent).
- TogglesEntities trait → dashboard, devices index, rooms show all drive the
same toggle (H1, audited); card grids replace the divide-y rows.
- Dashboard + devices grouped into responsive card grids by room.
- Windows page: sensor cards showing closed/open/tilted prominently, low-battery
chip, open sensors highlighted (screenshot-verified).
- Rooms index already card-based; left as-is.
Suite 51 green, 12/12 tabs clean, zero console errors.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Addresses "I have to configure MQTT per device; HA just works":
- Shared device account `shelly` (one credential every Shelly uses) with a
SCOPED ACL (status/events/online/rpc for any prefix; cannot touch homeos/,
ring/ or $SYS). gen-passwd seeds it from MQTT_SHELLY_PASSWORD.
- Auto-onboarding: IngestShellyMessage creates the device on the first
recognizable component (sys/wifi/cloud noise ignored), so pointing a Shelly
at the broker is all it takes — no manual "Zuweisen". Partial unique index on
config->>'mqtt_prefix' + race-safe create (merge-dedup migration).
- Settings → Geräte-MQTT card: server, username, reveal/copy password + steps
(config/homeos.php, MQTT_DEVICE_HOST). Discovery "Zuweisen" now upserts by
prefix (names/rooms an already-onboarded device, no duplicate) and no longer
forces per-device creds.
- Per-device credentials kept as opt-in hardening: a "generate" button on the
device page (pattern %u ACL retained).
Live-verified: publishing as `shelly` to a new prefix auto-creates the device
with switch+power state; a sys topic creates nothing. Suite 46 green, 12/12
tabs clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Second half of the Ring integration (handoff §12):
- RingTopics + RingNormalizer (H3): parse ring/<loc>/<cat>/<id>/<entity>/state,
map ding/motion/contact/lock/battery; bridge status topic drives the addon
connection state. Defensive — unknown topics are ignored, never junk devices.
- IngestRingMessage (H4, single ingest path): auto-creates cloud-flagged Ring
devices on first sight, monotonic race-safe state upsert, broadcasts live.
Gated on the addon being installed (cached) so a still-running bridge can't
recreate devices after uninstall.
- Shared AppliesDeviceState trait: the monotonic upsert now lives once, used by
both Shelly and Ring ingest (MqttTest guards the Shelly path).
- Listener subscribes ring/#. ring-mqtt container (opt-in `addons` compose
profile) as least-privileged `ring` MQTT user (ACL: ring/# only); gen-passwd
seeds the account. "Cloud" badge on Ring devices in list + detail.
11 Ring tests (topic parse, normalizer, bridge status, auto-create, install
gate, unknown-topic guard, out-of-order). Suite 40 green; 12/12 tabs clean.
Live-verified: real MQTT ring/.../ding + info publish → cloud device created
with ding + battery state. (Real Ring OAuth login runs in the ring-mqtt sidecar.)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
First half of the Ring integration (handoff §12: ring-mqtt, cloud token auth):
- Addons registry (static catalogue) + Addon model with encrypted config, so
the Ring refresh token / credentials are never stored in plaintext.
- AddonService: install / uninstall (wipes secrets) / saveConnection (blank
password keeps the stored one) / markStatus (from the bridge over MQTT).
- Addons page (new sidebar tab) with per-addon install, Cloud badge, live
status pill, and a Ring connect modal capturing email/password/2FA — we
never call Ring ourselves; creds are handed to the ring-mqtt sidecar.
- Uninstall behind a wire-elements confirm (R5). Fixed the confirm target:
Livewire names App\Livewire\Addons\Index as "addons" (Index suffix dropped),
and only $wire.$dispatch (not Alpine $dispatch) reaches the modal listener.
- DE/EN localization (R16); doorbell/package/cloud icons.
7 AddonTest cases (registry, install, encryption-at-rest, blank-password
keep, uninstall wipe, status guard). Suite 33 green; 12/12 tabs clean; full
install→connect→connecting→uninstall UI flow browser-verified, 0 console errors.
Backend bridge (ring-mqtt container) + Ring MQTT normalizer land next.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
New "Steuerung" page for tablet control:
- Large touch tiles for every light/switch — tap toggles (through
DeviceCommandService), highlighted when on, live via Echo.
- Drag-to-reorder via SortableJS (handle per tile); order persisted to
entities.panel_sort.
- Lights get a colour/brightness modal → Light.Set (brightness + rgb) on the
Shelly, with preset swatches + a custom colour picker. Driver contract gains
setLight(); command service audits it (H1).
Nav check 11/11 tabs clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Assigning a Shelly now issues a real broker credential so the physical device can
connect (onboarding was incomplete before):
- MqttCredentialProvisioner writes a mosquitto-compatible PBKDF2-SHA512 ($7$)
password line (in PHP) for username = the device's topic prefix, and touches a
reload trigger. A small wrapper in the mosquitto container (docker/mosquitto/
config/entrypoint.sh) SIGHUPs mosquitto so it re-reads the passwd live — no
restart. Verified: a provisioned device authenticates and publishes to its own
prefix (bound by the pattern %u ACL).
- The credential is shown once on the device page after assignment (enter it into
the Shelly). passwd is app-owned + world-readable so the web request can write it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- metrics table + Metric model; metrics:sample command (scheduled every minute,
pruned to 24h) records MQTT throughput (a Redis counter the ingest jobs bump)
and host CPU load + memory% from /proc.
- Chart.js line-chart Alpine island (themed from CSS tokens, dual axis) on the
Host page shows MQTT/min, CPU load and memory %. wire:ignore so the 10s health
poll doesn't re-init it. Nav check 10/10 clean (0 console errors).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- AssignDevice derives the Shelly MQTT prefix from the mDNS service instance name
(the device id) instead of the slugified topic identifier (which carries the
_shelly._tcp suffix); non-Shelly devices get no prefix.
- IngestDiscoveryMessage broadcasts DeviceDiscovered only on first creation, not on
every retained/periodic re-publish.
- The generic Confirm modal can target its event at the opening component (->to),
so a shared event name can't be caught by an unrelated component.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- UnifiClient service (art-of-wifi/unifi-api-client) talks to the local UDM with a
read-only account (self-signed cert, verify_ssl off). Verified live: logged in
and read 32 active clients.
- presence:poll command (scheduled every minute, withoutOverlapping): "home"
immediately on association, "away" only after an 8-min debounce so iPhone WLAN
sleep can't cause false-aways. Broadcasts PresenceChanged on the presence channel.
- Persons page: "Person hinzufügen" modal picks the representing device straight
from the live UniFi client list (falls back to manual MAC entry if UniFi is down);
the sweep then tracks that person's presence. mac + last_seen_home_at on persons.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Python discovery sidecar (zeroconf mDNS over host network) publishes findings
to homeos/discovery/<source>/<id> as the sidecar user. Compose service with
network_mode: host + NET_RAW. Verified live: it found real devices on the LAN
(a printer and Shellys) via mDNS.
- Listener also subscribes homeos/discovery/#; IngestDiscoveryMessage upserts
discovery_findings (preserving assigned/ignored) and broadcasts DeviceDiscovered
on the private discovery channel.
- "Neue Geräte" page lists findings live with Assign (modal → creates a Device
and links the finding) and Ignore/Restore. Per-device broker credentials are
provisioned at onboarding (noted in the assign hint).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Persons & Anwesenheit, Netzwerk & Discovery, Automationen, Einstellungen and
Zugang & Face-ID are now working Livewire pages (were dead "In Kürze" buttons).
Adds persons / discovery_findings / automations tables + models; authorizes the
presence + discovery broadcast channels; silences MQTT debug-log spam. Nav check
10/10 tabs clean (200, 0 console errors, 0 failed requests).
Deeper behaviour (presence polling, discovery sidecar, automation engine) lands
in the respective phase; the pages + schema are in place and render live.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Activates the windows nav tab. Summary banner (how many open) + per-room sensor
list with open/closed state and battery, live via Echo. Nav check 5/5 clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Activates the Räume nav tab. Index lists rooms with online/lights/open counts;
show lists a room's devices with entity chips and switch/light toggles (through
DeviceCommandService), live via Echo. Zero console errors (nav check: 4/4 clean).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Sidebar drawer is now opaque (.sidebar-tint solid), so on mobile the bright
content no longer bleeds through and it stays readable.
- Devices index reworked for ~50 devices: live search (name/model/vendor),
room + online/offline filters (URL-persisted), grouped by room with counts —
scannable instead of one long list.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Real bidirectional MQTT so devices are live, not mock (handoff §13.3):
- Mosquitto 2 broker (auth + per-client ACLs for laravel/shelly/sidecar from day
one); passwd generated by docker/mosquitto/gen-passwd.sh (gitignored).
- mqtt-listener daemon: subscribes `+/status/#`, parse + dispatch only (H2),
exponential reconnect backoff, graceful SIGTERM. php-mqtt/laravel-client.
- Ingest path (H4): IngestShellyMessage resolves device by mqtt_prefix, upserts
device_states, refreshes last_seen, broadcasts DeviceStateChanged
(ShouldBroadcastNow) on the private `home` channel.
- Control path (H1): DeviceDriver contract + ShellyMqttDriver (command topic +
Shelly.Reboot RPC) behind DeviceCommandService, which audits every command to
the new `commands` table. Device detail toggles + restart route through it;
flash reflects the real result.
- Live UI: dashboard + device pages listen via Echo (#[On('echo-private:home,
.DeviceStateChanged')]) and re-render instantly.
- Vendor specifics isolated in Support/Mqtt + Support/Drivers (H3).
Verified end-to-end in a real browser: publishing an MQTT status turned a light
"An" on the dashboard in 3.0s with no reload, 0 console errors. R12 30/30;
15 feature tests green (incl. ingest + command audit). README/bootstrap document
the broker passwd step.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
HomeStatus aggregated from Room, so a device with room_id = null (a supported
state, e.g. after picking "no room") vanished from dashboard totals, KPIs and
warnings. Aggregate from all devices instead; the dashboard groups them by room
with a trailing "Ohne Raum" group so nothing disappears. Added a feature test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Warnings no longer clutter the dashboard: a compact "Warnungen (n)" button in
the topbar opens a themed wire-elements/modal (R5) listing all messages. Dark
modal wrapper overridden; HomeStatus service shared by dashboard + modal.
- Device detail page (/devices/{uuid}, UUID route key): edit name, room and
active state; view info + live capabilities; Neustart (with confirm modal) and
"Update prüfen" as mock commands (Phase 3 routes them through the real driver).
Devices index (/devices) added; "Geräte" nav activated; dashboard device rows
and index link to the detail. Generic Confirm modal + x-detail component.
- Mock devices no longer rot: online is now "active and (no last_seen or seen
<10min)", so the demo stays healthy; the one offline device keeps a stale
timestamp. Full DE/EN i18n for devices + modal copy.
Verified: R12 30/30 in headless Chromium (0 console errors, 0 failed requests);
10 feature tests green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reworks the dashboard into a home-control view per user feedback (it is not a
server dashboard):
- Dashboard now shows home data — rooms with live device state (lights on/off,
power draw, window/door contacts, battery %), a home summary (devices online,
lights on, open contacts, low batteries), and a Warnings panel that only lists
what needs attention (offline device, open window, low battery, degraded host
service). No server internals on the dashboard.
- Server/service health (DB, Redis, Reverb, Horizon) + version info moved to a
dedicated "Host & Dienste" page (new nav item under System); dashboard surfaces
a host problem only as a warning that links there.
- New domain slice (handoff §3, mock-first §13.2): rooms/devices/entities/
device_states migrations + models (UUID route keys, R11) + DemoHomeSeeder with
Shelly-like devices incl. deliberate faults. Extracted SystemHealth service.
- Sidebar decluttered further; new x-entity-state component; full DE/EN i18n.
- Fixed R15 findings: .env.example now ships matching non-empty dev defaults for
DB_PASSWORD and Reverb keys so a fresh `cp .env.example .env` boots cleanly.
Verified: R12 21/21 in headless Chromium (0 console errors, 0 failed requests,
breakpoints 375/768/1280); 10 feature tests green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Addresses user feedback on Phase 1:
- Dashboard reoriented from static stack facts to LIVE operational status.
New health banner surfaces faults prominently (green when all online;
amber/red with a pluralized problem count when a service is down).
Service tiles probe DB, Redis, Reverb (socket) and Horizon in real time
with latency; versions moved to a small, de-emphasized "System" block.
wire:poll.10s + a refresh button keep it live. Verified: stopping Reverb
flips the banner to "Ein Dienst mit Störung" and the Echtzeit tile to
Offline/nicht erreichbar (screenshotted), healthy state restores clean.
- Sidebar decluttered: removed the wrapping "In Kürze" count badges;
coming-soon items are now dimmed (still show lock icons where gated).
R12 re-verified in headless Chromium: 15/15 assertions, 0 console errors,
0 failed requests, breakpoints 375/768/1280. 7 feature tests still green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
HomeOS Bootstrap