No description
Update tracking files to reflect completed implementation of: - 7-day pattern learning algorithm with Welford's online algorithm - Anomaly scoring against learned patterns (time score, zone score, composite) - Security mode integration (Armed, Disarmed, ArmedStay states) - Alert chain with timers (dashboard → webhook → escalation) - REST API endpoints for security mode and anomaly management - Database persistence for patterns and anomaly events - Cooldown deduplication and outlier protection All acceptance criteria met: - System detects deviations from learned patterns via multiple anomaly types - Accuracy improves measurably through feedback loop integration Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> |
||
|---|---|---|
| .beads | ||
| .github/workflows | ||
| .marathon | ||
| dashboard | ||
| docs | ||
| firmware | ||
| mothership | ||
| tests/e2e | ||
| .dockerignore | ||
| .gitignore | ||
| .needle-predispatch-sha | ||
| .needle.yaml | ||
| docker-compose.yml | ||
| Dockerfile | ||
| PROGRESS.md | ||
| VERSION | ||