TLS front door (Caddy) for this repo's own services: one hostname, a permanent
HTTP->HTTPS redirect, and the API on the same origin as the page. Home Assistant,
Grocy, Frigate and the rest keep their own ports — fronting HA brings its own
auth and websocket concerns and none of the problems below need it.
Three concrete reasons, not hygiene:
1. The admin panel's URL carries IDENTITY_TOKEN, which grants administrative
access to the person registry AND to the device grants that decide whether a
smart lock opens. On plain HTTP that is readable by anything on the smart-home
VLAN — a segment deliberately full of cheap IoT hardware.
2. getUserMedia requires a secure context, so register.html's camera cannot work
over http://192.168.x.x at all. It also failed SILENTLY: the call site used
`navigator.mediaDevices?.getUserMedia(...).then().catch()`, and optional
chaining short-circuits the whole chain — so neither handler ran, no "Camera
unavailable" message appeared, and registration proceeded photo-less with no
explanation. Verified in node. register.js now checks explicitly and says
which of the two cases it is; HTTPS is what actually fixes it.
3. Serving the page over HTTPS while ?api= still pointed at http://...:8097 would
have every call blocked as mixed content, so the API has to be proxied too.
tls: internal runs Caddy's own CA (no external dependency, works with no WAN at
all) with tools/export-proxy-ca.sh to fetch the root; tls: custom takes an
existing cert, which is how you'd use a real one from a DNS-01 challenge without
exposing anything. HSTS is deliberately not set — with an internal CA it would
turn a dismissible warning into a hard failure. Kiosks stay on plain HTTP for
now: a full-screen cert interstitial is not dismissible on a device with no
keyboard, so moving them is documented as a follow-up rather than done blind.
Empty service tokens now fill themselves in on the first build and land in
tokens.txt with what each is for. They are written BACK to the config, which is
the part that matters: a token is only useful because two machines agree on it,
so generating fresh randomness per build would produce a door panel that cannot
talk to the service it was built for. Blanks are filled once and never
overwritten. ha_token, mqtt_password, admin_password_hash and ssh_authorized_key
are deliberately not invented — tokens.txt lists them with the reason, so an
empty field is never a mystery.
32 new checks: token generation and stability across runs, disabled services
skipped, tokens.txt contents, config still valid after the rewrite, Caddyfile
routes and redirect, conditional pantry route, both TLS modes, and that the
derived admin URL keeps page and API on one origin. Nothing has been run against
a real Caddy — see proxy/README.md.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Two unchecked items from the README status list that were buildable in-repo
rather than blocked on hardware.
hosts/llm-host/ — Phase 3's LLM machine:
- Auto-detects gpu vs cpu tier (nvidia-smi must both exist AND succeed; a
leftover driver package on a machine whose card was pulled satisfies only
the first and would fail later at container start).
- Runs Ollama as a pinned container rather than curl|sh into a root shell,
matching how everything else here is deployed. Deliberately does NOT install
the GPU driver — the most hardware/kernel-specific step on that box.
- Sets OLLAMA_HOST=0.0.0.0 inside the container. Ollama binds loopback by
default, which in Docker means the published port forwards to nothing and
every caller sees a connection refused indistinguishable from "the host is
off" — and since every consumer here is built to tolerate exactly that, it
degrades silently. Same class of bug as chores' 127.0.0.1 env values.
- Takes a position on Ollama contention (open decision #4's resource half):
MAX_LOADED_MODELS=1 so a 14B text model and a vision model swap predictably
instead of thrashing VRAM or OOM-ing mid-request, NUM_PARALLEL=1 for
predictable Assist latency, KEEP_ALIVE=30m so a household that talks to
Assist a few times an hour isn't paying model-load cost every time.
- Documents that Ollama has NO authentication and its API can delete models,
not just generate — added to network-integration.md's port table, since the
network is the entire boundary.
docs/caldav-integration.md — Phase 8's notes:
- The four independent clients and their directions (digest-engine read-only,
chores' busy-check read-only, trash-calendar create-only under a UID-prefix
ownership invariant, HA's own bridge).
- Why they share one Nextcloud app password, and the two costs: rotation
touches three env files plus HA and fails quietly, and the read-only
invariant is a CODE property, not a permission boundary — an app password
can't be scoped read-only or per-calendar, so the server would not catch a
regression that started writing.
- The two traps worth knowing before debugging them: unexpanded recurrence
reporting a meeting on the day it was created, and CALDAV_VERIFY_TLS=false.
Neither has been run — no Debian machine, no GPU, no live Nextcloud. The script
is syntax-checked and its generated compose validated as YAML for both tiers;
that is the whole of the testing, and both READMEs say so.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Records the networking decision behind Phase 6b's arrival notifications so it
doesn't get re-derived later.
ntfy stays LAN-only: identity -> ntfy is a container-to-container call that never
reaches OPNsense, and remote delivery rides a WireGuard split tunnel routing only
the smart-home VLAN. No DMZ, no port forward, no certificates, no inter-VLAN rule.
Both alternatives were weighed and rejected, with reasons recorded:
- DMZ + port forward buys only "notifications without the VPN connected," at the
cost of a public listener, TLS upkeep, and mandatory ntfy auth — an exposed
topic leaks "the house is empty right now".
- DMZ + NAT reflection does avoid a new inter-VLAN rule (the smart-home VLAN's
existing outbound-WAN allow already covers traffic to your own public IP), but
makes a purely internal path depend on the ISP and external DNS being up, so
two devices metres apart stop talking during a WAN outage. CGNAT breaks it
outright, and the rule it avoids was the safe direction anyway.
New §2.1 covers the split tunnel itself, including the trap that motivated it:
AllowedIPs must be the actual VLAN prefix, never 192.168.0.0/16, which collides
with typical café/hotel LANs and breaks the phone's connectivity there. Plus
always-on VPN yes / lockdown no, and no DNS= so queries aren't pulled home.
The household uses no Apple devices, which removes the only forcing function for
exposure (ntfy's iOS app needs APNs and therefore an upstream relay). Open
decision #37 closes as a household decision; the Apple case is kept as a
footnote in case that ever changes.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Phase 19: trash-calendar (Kennelbach ICS -> CalDAV) and transit (GTFS
departures + OpenTripPlanner-backed voice route planning, slow walking
speed assumed). Phase 20: Tapo camera integration via identity's new
Frigate face-recognition presence signal, and chores, a presence/
calendar-driven household task nudger ("I don't care who does it, as
long as it gets done") with a passive fairness tally, litter culprit
attribution, and per-person chore_exempt/chore_reminder_style settings
on identity.
Also: the HA integrations + hardware-monitoring catalogs, Music
Assistant wiring, docs/network-integration.md (OPNsense VLAN
segmentation, no WAN port-forwards), the ESPHome BLE-proxy firmware,
and RuView CSI-presence integration (github.com/ruvnet/ruview) with
household automations for sleep/distress/concurrent-vitals/bathroom-
occupancy — all flagged with their real unverified assumptions and
open decisions. Fixed two real port collisions surfaced while wiring
this in (OTP vs. zigbee2mqtt on 8080, Music Assistant vs. pantry-vision
on 8095).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01K6CrKjW3yVptUnjG35HjC1
Amir Alexander Abdelbaki