config.yaml.example

# ============================================================
# BLE Scale Sync, Example Configuration
# ============================================================
# Copy this file to config.yaml and edit the values.
# Run "npm run validate" to check your config.
# Run "npm run setup" for an interactive wizard.
#
# Secrets: use ${ENV_VAR} references (resolved from .env or
# environment variables at load time).
# ============================================================

version: 1

# --- BLE Settings (optional, uncomment to pin to a specific scale) ---
# ble:
#   scale_mac: "FF:03:00:13:A1:04"   # Run "npm run scan" to find yours
#   bind_key: "${XIAOMI_BIND_KEY}"     # Xiaomi S800 / S400: 32-hex MiBeacon key from the Mi cloud
#   noble_driver: abandonware         # "abandonware" | "stoprocent"
#   handler: auto                     # "auto" | "mqtt-proxy" | "esphome-proxy" | "ha-bluetooth"
#   adapter: hci0                     # Linux only: select BLE adapter (omit for default)
#   force_scale_adapter: "Hutbit"     # override protocol auto-detection (requires scale_mac)
#   session_timeout_sec: 20           # seconds one GATT session waits for a reading (default 120)
#   qn_protocol_byte: 0               # QN scales only: 0 to 255, see below
#   qn_report_byte: 252               # QN scales only: 0 to 255, see below
#
# scale_mac: omit for auto-discovery. On macOS, use the CoreBluetooth UUID.
# noble_driver: default is node-ble (Linux), abandonware (Windows), stoprocent (macOS).
# handler: "auto" uses local Bluetooth. "mqtt-proxy" uses an ESP32 BLE-to-MQTT bridge.
#          "esphome-proxy" uses an existing ESPHome BT proxy via Native API (port 6053).
#          "ha-bluetooth" subscribes to Home Assistant's Bluetooth advertisement stream (broadcast scales only).
# adapter: Linux only, select which Bluetooth adapter to use (e.g., hci1 for a USB dongle).
#          Run "hciconfig" or "btmgmt info" to list available adapters.
#          Ignored on macOS/Windows. Omit to use the first adapter (hci0).
# force_scale_adapter: name of the scale protocol adapter to use, exactly as printed in the
#          "Adapters:" line at startup. Only for when auto-detection picks the wrong one;
#          it bypasses matching entirely, so scale_mac is required. Please also open an
#          issue when you need this, so the matcher itself can be fixed.
# session_timeout_sec: 5 to 600, native BLE handlers only. Lower it for scales that refuse to
#          run a standalone weigh-in while a host holds the session open (e.g. Beurer BF500,
#          which displays "APP"). Shorter sessions mean more adapter resets and trip the
#          runtime.watchdog_max_consecutive_failures limit sooner.
# qn_protocol_byte: QN-family scales only (Renpho, Arboleaf, FITINDEX, GE and rebadges). The
#          protocol byte the handshake echoes back to the scale. Set it only when a QN scale
#          connects, completes the whole handshake in the debug log and then never reports a
#          weight. Try 0, or 255, or the value the "QN: scale info" log line reports.
# qn_report_byte: QN-family scales only, and only worth trying after qn_protocol_byte did not
#          help. Payload byte of the history-response frame, default 254 (0xFE). Vendor-app
#          captures of a GE CS 10 G and an Arboleaf QN-Scale both send 252 (0xFC) instead.
#
# --- ESP32 MQTT Proxy (when handler: mqtt-proxy) ---
#   mqtt_proxy:
#     broker_url: "mqtt://192.168.1.50:1883"  # MQTT broker URL
#     device_id: esp32-ble-proxy               # must match ESP32 config.json
#     topic_prefix: ble-proxy                  # must match ESP32 config.json
#     username: myuser                         # optional
#     password: "${MQTT_PASSWORD}"             # optional
#
# --- Home Assistant Bluetooth (when handler: ha-bluetooth) ---
#   ha_bluetooth:
#     url: "http://homeassistant.local:8123"  # HA base URL (http/https) or ws(s)://.../api/websocket
#     token: "${HA_TOKEN}"                    # long-lived access token of an ADMIN user, from .env
#     source: "9c:13:9e:34:82:08"           # optional: only this HA scanner (source id)

# --- Scale Units ---

scale:
  weight_unit: kg   # kg | lbs (display only, calculations always use kg)
  height_unit: cm   # cm | in  (for height input below)

# --- Multi-User Matching ---
# Strategy when weight falls outside all configured ranges:
#   "nearest" = assign to closest range midpoint (with warning)
#   "log"     = log warning and skip the measurement
#   "ignore"  = silently skip the measurement
unknown_user: nearest

# What to do with a reading no user's weight_range covers.
#   "warn" (default) = log it and export it anyway (the behaviour before
#                      this option existed)
#   "skip"           = log it and stop: no export, and last_known_weight is
#                      left alone so the next weigh-in still matches correctly
out_of_range: warn

# --- Users ---
# At least one user is required. For multi-user setups,
# define weight_range so the app can identify who stepped on the scale.

users:
  - name: Alice
    slug: alice                          # unique, lowercase, used in MQTT topics & InfluxDB tags
    height: 168
    birth_date: "1995-03-20"
    gender: female
    is_athlete: false
    weight_range: { min: 50, max: 75 }   # always in kg
    last_known_weight: null               # auto-updated after each measurement

    # Beurer BF7xx / BF9xx only (consent-gated SIG scales):
    # beurer_pin: 1234                    # code the scale was paired with; 0 on a wiped scale
    # beurer_user_index: 1                # scale user slot the code belongs to
    # beurer_provision: false             # write this profile into a scale that has none stored

    # Per-user exporters (optional, override global_exporters for this user):
    # exporters:
    #   - type: garmin
    #     email: alice@example.com
    #     password: "${GARMIN_PASSWORD_ALICE}"
    #     token_dir: ./garmin-tokens/alice
    #   - type: strava
    #     client_id: "${STRAVA_CLIENT_ID}"
    #     client_secret: "${STRAVA_CLIENT_SECRET}"
    #     token_dir: ./strava-tokens/alice   # run "npm run setup-strava" after config
    #   - type: intervals
    #     athlete_id: i123456                # Intervals.icu Settings → Developer
    #     api_key: "${INTERVALS_API_KEY}"

  - name: Bob
    slug: bob
    height: 183
    birth_date: "1990-06-15"
    gender: male
    is_athlete: true
    weight_range: { min: 75, max: 100 }
    last_known_weight: 85.5

# --- Global Exporters ---
# Shared by all users (unless a user defines their own "exporters" list).
# All enabled exporters run in parallel.

global_exporters:
  # -- Garmin Connect --
  - type: garmin
    email: "${GARMIN_EMAIL}"
    password: "${GARMIN_PASSWORD}"
    token_dir: ~/.garmin_tokens
    # weight_only: true    # upload the weight alone; BMI, body fat, water, bone,
    #                      # muscle, visceral fat, physique rating and metabolic
    #                      # age are left unset in Garmin Connect

  # -- MQTT (Home Assistant auto-discovery) --
  # - type: mqtt
  #   broker_url: "mqtt://localhost:1883"  # use mqtts:// for TLS
  #   topic: scale/body-composition
  #   qos: 1                               # 0, 1, or 2
  #   retain: true
  #   username: myuser
  #   password: "${MQTT_PASSWORD}"
  #   client_id: ble-scale-sync
  #   ha_discovery: true                   # Home Assistant auto-discovery
  #   ha_device_name: BLE Scale

  # -- Webhook (generic HTTP) --
  # - type: webhook
  #   url: "https://example.com/hook"
  #   method: POST
  #   headers:
  #     X-Api-Key: secret123
  #   timeout: 10000

  # -- InfluxDB v2 --
  # - type: influxdb
  #   url: "http://localhost:8086"
  #   token: "${INFLUXDB_TOKEN}"
  #   org: my-org
  #   bucket: my-bucket
  #   measurement: body_composition

  # -- Ntfy (push notifications) --
  # - type: ntfy
  #   url: "https://ntfy.sh"
  #   topic: my-scale
  #   title: Scale Measurement
  #   priority: 3
  #   token: "${NTFY_TOKEN}"               # Bearer token auth
  #   # username: myuser                   # or Basic auth
  #   # password: "${NTFY_PASSWORD}"

  # -- Telegram (bot notifications) --
  # - type: telegram
  #   bot_token: "${TELEGRAM_BOT_TOKEN}"   # token from @BotFather
  #   chat_id: "987654321"                 # numeric ID or @channelusername
  #   title: Scale Measurement
  #   silent: false                        # true = no notification sound

  # -- File (CSV/JSONL) --
  # - type: file
  #   file_path: "./measurements.csv"      # path to output file
  #   format: csv                          # "csv" or "jsonl"

  # -- Wger (self-hosted weight/workout manager) --
  # - type: wger
  #   base_url: https://wger.de            # your instance URL
  #   token: "${WGER_TOKEN}"               # API key from <base_url>/en/user/api-key
  #   sync_measurements: true              # also push body fat/water/muscle/bone

# --- Runtime Settings ---

runtime:
  continuous_mode: false    # true = keep scanning in a loop (ideal for Raspberry Pi)
  scan_cooldown: 30         # seconds between scans in continuous mode (5–3600)
  idle_rescan_delay: 5      # seconds before rescanning when no scale was found (0-3600)
  retry_failed_exports: true # keep a reading whose export failed and retry it later
  dry_run: false            # true = read scale + compute body comp, skip all exports
  debug: false              # true = verbose BLE logging

# --- Update Check ---
# On each successful measurement (max once per 24h), the app sends a lightweight
# GET request to check for newer versions. Only the app version, OS, and architecture
# are sent via the User-Agent header. No personal data is collected.
# Disable with: update_check: false
update_check: true