Preview what a source-priority resolution of a day's raw measurements would look like, alongside daily_metrics' actual (all-sources) current value for that day.

Assess the severity and context of an anomalous value against recent baseline.

Assess the strength and reliability of a cross-metric correlation.

Assess the statistical and personal significance of a metric trend, using baseline confidence and direction interpretation.

Assess the significance of a period-over-period change (e.g. this week vs. last week).

Compute mean, median, and standard deviation of a metric over a date range for longitudinal comparison.

Calculate direction and slope of a metric's trend over time using ordinary least squares regression against calendar days.

Clear (blank out) a single daily metric for a given day, undoing a previous log_daily_metric call.

Compare two non-overlapping periods of the same metric (e.g. this week vs. prior 4 weeks).

Flag values that deviate sharply from a metric's own baseline using modified z-score (median + MAD), robust to outliers in small samples.

Export health data to a CSV file for external analysis or backup.

Compute Pearson correlation between two metrics, joined by date, with optional lag for predictive relationships.

Break a metric's readings for a day down by source, to spot when two sources disagree on the same day.

Log one or more daily metrics (steps, sleep hours, heart rate, weight, workout minutes, mood, water) for a specific day.

Record a single raw, timestamped observation (with optional source/unit) instead of a whole day's summary.

Read daily health metrics (steps, sleep, heart rate, weight, workouts, mood, water) over a date range, with optional filtering and aggregation.

Read raw measurement rows, filterable by metric/date range/source. Returns timestamped observations with their source and unit.