feat(insights): inject guideline band placement into generic metric assessments

Thread the interpretation block into the generic per-metric assessment
path (the ~40 HealthKit metric cards). For a metric the knowledge base
covers, the prompt now carries the band table plus the server-computed
position of the current value, so the assessment names where the value
sits on the guideline scale and judges the trend by that position — the
visceral-fat 2.7 case now reads as "comfortably in the healthy band"
rather than reciting how many times it was logged.

The block is absent for uncovered metrics and for a sex-split band with
no known sex, so those cards fall back to exactly today's personal-relative
behaviour. Band edges carry no strip risk here — the generic assessment
path runs no post-generation numeric grounding gate.

src/lib/ai/prompts/__tests__/assessment-prompt-contract.test.ts

@@ -119,6 +119,25 @@ describe("metric-archetype user prompt — context block plumbing", () => {
     expect(out).not.toContain("undefined");
   });
 
+  it("appends the interpretation block when provided (Welle J)", () => {
+    const out = getMetricArchetypeUserPrompt(
+      meta,
+      "{}",
+      "2026-06-05",
+      "en",
+      "",
+      "",
+      "INTERPRETATION CONTEXT (guideline bands): the healthy band.",
+    );
+    expect(out).toContain("INTERPRETATION CONTEXT");
+  });
+
+  it("omits the interpretation block cleanly when absent", () => {
+    const out = getMetricArchetypeUserPrompt(meta, "{}", "2026-06-05", "en");
+    expect(out).not.toContain("INTERPRETATION CONTEXT");
+    expect(out).not.toContain("undefined");
+  });
+
   it("makes the closing step conditional in the user instruction (both locales)", () => {
     expect(
       getMetricArchetypeUserPrompt(meta, "{}", "2026-06-05", "en"),

src/lib/ai/prompts/metric-archetypes.ts

@@ -127,6 +127,13 @@ export function getMetricArchetypeUserPrompt(
    * grounded in already-computed data; optional and may be empty.
    */
   assessmentContextBlock?: string,
+  /**
+   * v1.27.13 (Welle J) — the guideline interpretation block (band table +
+   * server-computed position) for metrics the knowledge base covers. Absent
+   * for metrics with no guideline band — the prompt then stays personal-
+   * relative, exactly as before.
+   */
+  interpretationBlock?: string,
 ): string {
   const ctxBlock =
     previousContextBlock && previousContextBlock.trim().length > 0
@@ -136,14 +143,18 @@ export function getMetricArchetypeUserPrompt(
     assessmentContextBlock && assessmentContextBlock.trim().length > 0
       ? `\n\n${assessmentContextBlock}\n`
       : "";
+  const interpBlock =
+    interpretationBlock && interpretationBlock.trim().length > 0
+      ? `\n\n${interpretationBlock}\n`
+      : "";
   if (locale === "en") {
     return `Date: ${todayKey} (Europe/Berlin)
-Write one short assessment of this person's ${meta.displayName.toLowerCase()}: name the current level with a concrete number from the snapshot, place the recent window against their own weekly/monthly baseline, and — when something is genuinely actionable — close with one doable step; when nothing is, skip the step rather than manufacture filler. Judge confidence from the measurement count and recency.${ctxBlock}${extraBlock}
+Write one short assessment of this person's ${meta.displayName.toLowerCase()}: name the current level with a concrete number from the snapshot, place the recent window against their own weekly/monthly baseline, and — when something is genuinely actionable — close with one doable step; when nothing is, skip the step rather than manufacture filler. Judge confidence from the measurement count and recency.${ctxBlock}${extraBlock}${interpBlock}
 
 ${snapshotJson}`;
   }
   return `Datum: ${todayKey} (Europe/Berlin)
-Schreibe eine kurze Einschätzung zu ${meta.displayName} dieser Person: benenne das aktuelle Niveau mit einer konkreten Zahl aus dem Snapshot, ordne das recent-Fenster gegen die eigene Wochen-/Monats-Baseline ein und schließe — wenn etwas wirklich umsetzbar ist — mit einem machbaren Schritt; ist nichts umsetzbar, lass den Schritt weg statt Fülltext zu erfinden. Konfidenz aus Messanzahl und Aktualität ableiten.${ctxBlock}${extraBlock}
+Schreibe eine kurze Einschätzung zu ${meta.displayName} dieser Person: benenne das aktuelle Niveau mit einer konkreten Zahl aus dem Snapshot, ordne das recent-Fenster gegen die eigene Wochen-/Monats-Baseline ein und schließe — wenn etwas wirklich umsetzbar ist — mit einem machbaren Schritt; ist nichts umsetzbar, lass den Schritt weg statt Fülltext zu erfinden. Konfidenz aus Messanzahl und Aktualität ableiten.${ctxBlock}${extraBlock}${interpBlock}
 
 ${snapshotJson}`;
 }

src/lib/insights/metric-status.ts

@@ -26,6 +26,7 @@ import {
   getMetricArchetypeSystemPrompt,
   getMetricArchetypeUserPrompt,
 } from "@/lib/ai/prompts/metric-archetypes";
+import { buildInterpretationBlock } from "@/lib/ai/prompts/interpretation-block";
 import {
   getMetricStatusMeta,
   metricStatusScope,
@@ -461,6 +462,24 @@ export async function generateMetricStatus(args: {
     locale,
   );
 
+  // v1.27.13 (Welle J) — the guideline interpretation block for metrics the
+  // knowledge base covers (visceral fat, resting HR, SpO₂, respiratory rate,
+  // body temperature, sleep duration, waist, PWV). The band position is
+  // computed from the signal's current value (or the latest day when no signal
+  // window exists); absent for uncovered metrics (fail-soft to personal-
+  // relative). Sex-split bands (waist) resolve only when the profile carries a
+  // sex.
+  const interpretationValue = signal?.current ?? latest?.value ?? null;
+  const interpretationBlock =
+    interpretationValue !== null
+      ? buildInterpretationBlock({
+          metricKey: meta.id,
+          value: interpretationValue,
+          sex,
+          locale,
+        })
+      : undefined;
+
   const outcome = await runStatusCompletion({
     userId: args.userId,
     cacheAction,
@@ -473,6 +492,7 @@ export async function generateMetricStatus(args: {
       locale,
       previousContextBlock,
       contextBlock,
+      interpretationBlock,
     ),
     // v1.12.1 (D1) — the phrasing task benefits from a touch more sampling
     // entropy while the FACTS stay pinned by the snapshot + the