Analytics

An analytics plugin computes custom metrics, charts and insights for a single exercise over its whole history. The user picks one in Settings → Analytics; it renders on the exercise detail page.

The contract

export const pluginBundle = {
  formatVersion: 1,
  pluginId: "com.you.volume",
  family: "analytics",
  entryExport: "analytics",
} as const;

export const analytics = {
  id: "com.you.volume",
  name: "Weekly Volume",
  description: "Total tonnage per week.",
  metricDefinitions: [
    { id: "weeklyVolume", label: "Weekly volume", unit: "kg" },
  ],

  compute(input) {
    const byWeek = new Map();
    for (const entry of input.history) {
      const wk = Math.floor(entry.performedAtMs / (7 * 864e5));
      const vol = entry.sets.reduce((s, x) => s + x.weight * x.reps, 0);
      byWeek.set(wk, (byWeek.get(wk) ?? 0) + vol);
    }
    const points = [...byWeek.entries()]
      .sort((a, b) => a[0] - b[0])
      .map(([wk, v]) => ({ date: wk * 7 * 864e5, value: v }));

    return {
      metrics: [{ id: "weeklyVolume", value: points.at(-1)?.value ?? 0 }],
      series: [{ metricId: "weeklyVolume", label: "Weekly volume", points }],
      insights: points.length > 1 && points.at(-1).value > points.at(-2).value
        ? ["Volume up week over week"]
        : [],
    };
  },
};

Input & output

compute(input: AnalyticsInput) — @logit/core/domain/analytics.

• input.history — ExerciseHistoryEntry[], oldest first, all available.
• input.exercise — { name, id?, exerciseType? }.

type AnalyticsOutput = {
  metrics: { id: string; value: number | string; formatted?: string }[];
  series: { metricId: string; label: string; points: { date: number; value: number }[] }[];
  insights?: string[];
  label?: string;
};

metricDefinitions (id + label + optional unit / higherIsBetter) is read from a meta call at install and shown in the picker.

Nutrition analytics (nutrition-analytics) has the same shape over diary + weight data.