Dt Obs Predictive Analytics

This adds forecasting and anomaly detection to Dynatrace observability using DQL and the built-in analyzer tools. You get capacity planning with timeseries-forecast, trend onset detection with timeseries-novelty-detection, and anomaly flagging with adaptive and seasonal detectors. The documentation makes a sharp distinction worth understanding: novelty detection answers "did this metric change" while anomaly detectors answer "is this currently wrong." The skill pushes you toward the forecast analyzer for real predictions and away from naive linear extrapolation. Results come formatted as ranked tables with trend indicators and action flags. If you're trying to answer "which hosts hit 90% CPU in 30 days" or "when did memory usage start climbing," this gives you the query patterns and tool routing you need.