alee/vigilar
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
72 Commits 1 Branch 0 Tags 564 KiB
d69bf6d6af
Commit Graph

72 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Aaron D. Lee
4c9ebe029d Integrate YOLOv8 detection and pet ID into camera worker 2026-04-03 13:22:25 -04:00
Aaron D. Lee
d0acf7703c Handle pet and wildlife events in event processor 2026-04-03 13:20:46 -04:00
Aaron D. Lee
53daf58c78 Add camera_location filtering to alert profile matching 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:20:37 -04:00
Aaron D. Lee
e48ba305ea Add pet ID model trainer with MobileNetV3-Small transfer learning 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:19:12 -04:00
Aaron D. Lee
c7f9304f2a Add pet ID classifier with species-filtered identification 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:18:42 -04:00
Aaron D. Lee
13b7c2a219 Add wildlife threat classification with size heuristics 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:17:21 -04:00
Aaron D. Lee
131eed73b1 Add YOLOv8 unified detector with class classification 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:15:31 -04:00
Aaron D. Lee
d83710b839 Add pet and wildlife database query functions 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:13:56 -04:00
Aaron D. Lee
ea092bca10 Add pets, pet_sightings, wildlife_sightings, pet_training_images tables 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:11:01 -04:00
Aaron D. Lee
4ca06651b6 Fix import sort order in config.py (ruff I001) 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:08:59 -04:00
Aaron D. Lee
6e3ef1dcdc Add pet detection, wildlife, and activity config models 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:08:40 -04:00
Aaron D. Lee
aae857ec53 Add pet/wildlife enums, event types, and MQTT topics 
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
 2026-04-03 13:07:09 -04:00
Aaron D. Lee
767bbf26ba Add .worktrees/ and .superpowers/ to .gitignore 
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-03 13:02:10 -04:00
Aaron D. Lee
0c0f484cdf Add pet-aware security features implementation plan 
18 tasks covering: YOLOv8 detector, pet ID classifier, wildlife threat
classification, crop management, alert integration, web UI, and training.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-03 13:00:45 -04:00
Aaron D. Lee
4b8da811df Add pet-aware security features design spec 
Covers pet detection (YOLOv8), pet ID classifier, wildlife threat
tiers, zone-based alerting, training UI, and pet dashboard.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-03 12:50:14 -04:00
Aaron D. Lee
11d776faa6 Add smart alert profiles and recording timeline (Tasks 4-5) 
Task 4 — Alert Profiles: presence-aware + time-of-day alert routing.
Profiles match household state (EMPTY/KIDS_HOME/ADULTS_HOME/ALL_HOME)
and time windows (sleep hours). Per-detection-type rules control
push/record/quiet behavior with role-based recipients (all vs adults).

Task 5 — Recording Timeline: canvas-based 24h timeline per camera
with color-coded segments (person=red, vehicle=blue, motion=gray).
Click-to-play, date picker, detection type filters, hour markers.
Timeline API endpoint returns segments for a camera+date.

All 5 daily-use feature tasks complete. 140 tests passing.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-03 00:09:08 -04:00
Aaron D. Lee
8314a61815 Add presence detection, person/vehicle AI detection, health monitoring 
Task 1 — Presence: ping family phones, derive household state
(EMPTY/KIDS_HOME/ADULTS_HOME/ALL_HOME), configurable departure delay,
per-member roles, auto-arm actions via MQTT.

Task 2 — Detection: MobileNet-SSD v2 via OpenCV DNN for person/vehicle
classification. Vehicle color/size fingerprinting for known car matching.
Zone-based filtering per camera. Model download script.

Task 3 — Health: periodic disk/MQTT/subsystem checks, auto-prune oldest
non-starred recordings on disk pressure, daily digest builder.

126 tests passing.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-03 00:06:45 -04:00
Aaron D. Lee
8a65ac8c69 Add implementation plan for daily use features 
5-task plan covering presence detection, person/vehicle AI detection,
smart alert profiles, recording timeline UI, and health monitoring.
Tasks 1-3 parallelizable, 4 depends on 1+2, 5 depends on 2.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-03 00:01:53 -04:00
Aaron D. Lee
53ae925a70 Add daily use features design spec 
Spec covers 5 feature areas for making Vigilar a system a household
relies on daily: multi-person presence detection, MobileNet person +
vehicle detection with driveway fencing, smart alert profiles with
presence/time awareness, recording timeline UI, and health monitoring
with auto-prune and daily digest.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-02 23:42:29 -04:00
Aaron D. Lee
ebcc49b474 Add kiosk setup and deployment scripts (Phases 5 + 9) 
Phase 5 — RPi Kiosk:
- setup_kiosk.sh: full RPi OS Lite setup (X11, Chromium kiosk mode,
  auto-login, DPMS disabled, GPU memory split, screen rotation)
- kiosk.service: systemd unit for reliable auto-start
- update_kiosk.sh: reconfigure URL/rotation/resolution without re-setup
- Handles both Bullseye and Bookworm RPi OS versions

Phase 9 — Hardening + Deployment:
- install.sh: full server setup (apt/pacman, vigilar user, venv,
  directories, permissions, mosquitto config, systemd units)
- gen_cert.sh: TLS cert via mkcert or openssl fallback
- gen_vapid_keys.sh: VAPID keys for Web Push notifications
- setup_nut.sh: NUT configuration with USB UPS auto-detection
- backup.sh: SQLite snapshot + config archive, cron-ready
- uninstall.sh: clean removal with data preservation option
- vigilar.service: hardened systemd unit (ProtectSystem, NoNewPrivileges,
  PrivateTmp, syscall filtering)
- vigilar-mosquitto.conf: localhost-only MQTT broker config

All scripts idempotent, bash -n validated, support Debian + Arch.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-02 23:25:03 -04:00
Aaron D. Lee
10b0cf4d0e Add events/rules engine, sensor bridge, and UPS monitor (Phases 6-8) 
Phase 6 — Events + Rule Engine:
- EventProcessor subprocess: subscribes to all MQTT events, logs to DB,
  evaluates rules, fires alert actions
- ArmStateFSM: DISARMED/ARMED_HOME/ARMED_AWAY with PIN verification
  (HMAC-safe), DB persistence, MQTT state publishing
- RuleEngine: AND/OR logic, 4 condition types (arm_state, sensor_event,
  camera_motion, time_window), per-rule cooldown tracking
- SSE event stream with subscriber queue pattern and keepalive
- Event acknowledge endpoint

Phase 7 — Sensor Bridge:
- SensorBridge subprocess: subscribes to Zigbee2MQTT, normalizes payloads
  (contact, occupancy, temperature, humidity, battery, linkquality)
- GPIOHandler: conditional gpiozero import, callbacks for reed switches
  and PIR sensors
- SensorRegistry: maps Zigbee addresses and names to config sensor IDs
- SensorEvent/SensorState dataclasses
- Web UI now shows real sensor states from DB

Phase 8 — UPS Monitor:
- UPSMonitor subprocess: polls NUT via pynut2 with reconnect backoff
- State transition detection: OL→OB (power_loss), charge/runtime
  thresholds (low_battery, critical), OB→OL (restored)
- ShutdownSequence: ordered shutdown with configurable delay and command
- All conditionally imported (pynut2, gpiozero) for non-target platforms

Fixed test_db fixture to use isolated engines (no global singleton leak).
96 tests passing.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-02 23:17:53 -04:00
Aaron D. Lee
845a85d618 Initial commit: Vigilar DIY home security system 
Phase 1 (Foundation): project skeleton, TOML config + Pydantic validation,
MQTT bus wrapper, SQLite schema (9 tables), Click CLI, process supervisor.

Phase 2 (Camera): RTSP capture via OpenCV, MOG2 motion detection with
configurable sensitivity/zones, adaptive FPS recording (2fps idle/30fps
motion) via FFmpeg subprocess, HLS live streaming, pre-motion ring buffer.

Phase 3 (Web UI): Flask + Bootstrap 5 dark theme, 6 blueprints, Jinja2
templates (dashboard, kiosk 2x2 grid, events, sensors, recordings, settings),
PWA with service worker + Web Push, full admin settings UI with config
persistence.

Remote Access: WireGuard tunnel configs, nginx reverse proxy with HLS
caching + rate limiting, bandwidth-optimized remote HLS stream (426x240
@ 500kbps), DO droplet setup script, certbot TLS.

29 tests passing.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-02 23:11:27 -04:00