Files
internet_use_control/environment.d/control-service.env.example
Sam Rolfe 4f233a8012 Phase 8: overrides (shared keycode + versioned admin lock/unlock/grant API)
Build on Phases 1–7. Adds the two override paths from AGENTS.md §4 FR-10/FR-11:
the shared device keycode and the versioned admin HTTP API the separate Kotlin
admin app will call.

- store: migration 004 adds state.grant_until and a settings table. New
  overrides.go records every override in `overrides` and verifies the shared
  keycode (bcrypt; one code for every device, §10 rule 3). ExpireLocks also
  clears an expired grant, and NextScheduleBoundary includes the grant end so a
  device re-locks on time.
- effective state: a grant unlocks a device inside a schedule window; a manual
  lock clears a grant and a grant clears a manual lock. The retained config
  payload gains grant_until (additive, v stays 1).
- web: /api/v1/admin/... (list, state, lock, unlock, grant) guarded by
  CONTROL_ADMIN_API_TOKEN, separate from the session login, plus the on-device
  keycode endpoint POST /api/v1/device/override with per-address throttling.
  A state change is published retained before the API responds.
- cmd/control: `keycode` subcommand and start-up seeding from
  CONTROL_DEVICE_KEYCODE_HASH.
- deploy/verify-phase8.sh + `make verify-phase8` prove the definition of done:
  an API lock reaches a real broker subscriber in well under a second.
- Version 0.6.0-phase8 (compose, Makefile, docs).

Measured on .13: the lock reached the TLS broker subscriber in 7 ms (handler
1 ms, request to retained publish accepted).

Tests: go vet ./... and go test -count=1 ./... pass in golang:1.25-alpine.

Co-Authored-By: Paperclip <noreply@paperclip.ing>
2026-10-13 16:27:46 +11:00

90 lines
3.9 KiB
Plaintext

# Control Service configuration — template.
#
# Copy to control-service.env and fill in real values. The real file is
# gitignored and must never be committed. docker compose loads it with env_file.
#
# Secrets in this lab also live in ~/.config/environment.d/10-secrets.conf; the
# same variable names can be exported there instead. The service reads only the
# process environment, so either source works.
# Listen address inside the container.
CONTROL_ADDR=:8791
# SQLite database on the persistent volume.
CONTROL_DB_PATH=/data/control.db
CONTROL_APP_ENV=production
CONTROL_LOG_LEVEL=info
# Timezone weekly schedule windows are evaluated in. Windows are wall-clock, so
# this must be the family's zone, not the container's default of UTC.
CONTROL_TIMEZONE=Australia/Sydney
# Signs the session cookie. Set to a long random string (>= 16 chars).
# Generate: openssl rand -hex 32
CONTROL_SESSION_SECRET=change-me-to-a-long-random-string
# Session lifetime (Go duration).
CONTROL_SESSION_TTL=12h
# Secure cookie flag. Keep true behind Caddy/HTTPS. Set false only for
# plain-HTTP local testing.
CONTROL_COOKIE_SECURE=true
# The single dashboard login (Sam).
CONTROL_ADMIN_USERNAME=sam
# Option A (simple): plaintext password. Constant-time compared at login.
CONTROL_ADMIN_PASSWORD=change-me
# Option B (preferred): a bcrypt hash; leave CONTROL_ADMIN_PASSWORD empty.
# Generate: docker run --rm --entrypoint /app/control webrestrict/control:0.6.0-phase8 hash-password 'your-password'
# CONTROL_ADMIN_PASSWORD_HASH=$2a$10$...
# --- Admin API (Phase 8) ---------------------------------------------------
# The versioned HTTP API the separate Kotlin admin app (kotlin_phone_app_admin)
# calls directly. Its bearer token is SEPARATE from the dashboard session login.
# Generate a long random value: openssl rand -hex 32
# When empty the API answers 503 "admin API is not configured" and nothing else
# changes, so an older deployment keeps working.
CONTROL_ADMIN_API_TOKEN=change-me-to-a-long-random-token
# The shared device keycode (one code, every device — AGENTS.md §10 rule 3).
# Give a bcrypt hash, never the code itself. Generate one with:
# docker run --rm --entrypoint /app/control webrestrict/control:0.6.0-phase8 \
# hash-password 'the-code'
# The hash is seeded into SQLite at start-up and can be changed later with
# `docker exec webrestrict-control /app/control keycode`.
# CONTROL_DEVICE_KEYCODE_HASH=$2a$10$...
# --- MQTT (Phase 3) --------------------------------------------------------
# The broker is transport only. SQLite stays the source of truth, and only the
# Control Service publishes config (AGENTS.md §10 rules 5-6).
# Turn the broker link on or off. With it off the dashboard still works and the
# config stays in SQLite, unpublished.
CONTROL_MQTT_ENABLED=true
# The shared Mosquitto on `.13`, seen from inside this container. The compose
# file maps host.docker.internal to the host gateway. The TLS listener on 8883
# has been live since 2026-10-13 (deploy/mosquitto/README.md).
CONTROL_MQTT_BROKER=tls://host.docker.internal:8883
# The Control Service identity. The broker ACL gives it write access to every
# config topic and read access to every status and telemetry topic, nothing else.
# Generate the account with deploy/mosquitto/provision-clients.sh, which writes
# the values into deploy/mosquitto/clients.env (gitignored).
CONTROL_MQTT_CLIENT_ID=control-service
CONTROL_MQTT_USERNAME=control-service
CONTROL_MQTT_PASSWORD=change-me
# PEM bundle used to verify the broker certificate. Required for a tls:// broker.
# docker-compose.yml mounts deploy/mosquitto/tls/ here read-only.
CONTROL_MQTT_CA_FILE=/etc/webrestrict/tls/ca.crt
# Escape hatch for a broker that has no TLS listener: it allows a plaintext,
# credential-less link. The service refuses to start without credentials or TLS
# unless this is true, and it logs a warning when it is. Leave it false against
# the shared Mosquitto (AGENTS.md §10 rule 6).
CONTROL_MQTT_INSECURE=false