offline_sos_system 0.0.6

A pure Dart, offline-first Flutter package for on-device crash and anomaly detection using hardware telemetry and TFLite.

0.0.6

• Configurable Crash Detection: Introduced CrashDetectionConfig to allow custom thresholds for the sensor impact gate (impactThresholdG), minimum model confidence probability (minCrashProbability), fallback impact threshold (fallbackImpactThresholdG), and duplicate suppression cooldown (cooldown).
• Backward Compatibility: Preserved all existing default behaviors (3.0 G impact gate, 0.25 probability, 5.0 G fallback, 30s cooldown). Applications calling SafePulseEngine() without arguments continue to behave identically.
• Lifecycle & Reliability Improvements: Fixed engine disposal to properly cancel active sensor subscriptions and timers, guarded stream controllers against post-disposal emissions, prevented restarted sensor retries after stop, and ensured idempotent disposal.
• Comprehensive Automated Tests: Added full test coverage for configuration validation, exact threshold boundaries, offline fallback heuristics, duplicate suppression, and engine lifecycle.

0.0.5

• Fixed dangling library doc comment lint to ensure perfect pub.dev score.

0.0.4

• Added comprehensive Dartdoc comments to all public API elements for perfect pub.dev scoring.

0.0.3

• Massive documentation overhaul: Introduced advanced Edge AI architecture guides and high-conversion implementation examples.

0.0.2

• Initial release.
• Extracted the core offline detection engine into a standalone, pure Dart package.
• Integrated sensors_plus for continuous accelerometer and gyroscope telemetry.
• Integrated tflite_flutter for offline, on-device anomaly inference.
• Added support for optional customModelPath injection with a bundled default model fallback.
• Exposed Stream<CrashEvent> for decoupled app-layer integration.