detecting-anomalies

## What problem does it solve? This Skill helps teams detect anomalies in observability metrics, reducing mean time to detect by surfacing unusual patterns in time-series data.

## Core Features & Use Cases

• Statistical outlier detection (Z-score, IQR) to flag unusual latency, error counts, or throughput.
• Threshold and percentile-based alerts to enforce SLOs and capacity limits.
• Change-detection patterns (rate of change, moving-average deviation) for spikes and trend shifts.
• Noise-robust detection and multi-method consensus to reduce false positives.

### Quick Start

• Analyze the latest 60 minutes of your time-series metrics (latency, error rate, throughput) to surface anomalies using Z-score or rate-of-change calculations.
• Example: detect anomalies in latency with Z-score and moving-average deviation over the last hour.