feat(mood): revise factor + tag catalog (migration 0126)

Tighten the operator-finalised mood catalog to three RATED factors and a
trimmed default BINARY tag set.

- Seed `factor_sadness` (RATED, 1..5, inverse — higher = worse) under the
  `factors` category, with its label in all six locales and a CloudRain
  glyph. The RATED set is now work / sleep quality / sadness.
- Retire `factor_conflict`, `factor_family`, `factor_social`,
  `factor_stress`: flip each to BINARY and deactivate. The catalog rows
  stay so historical entries still resolve their labels (the insights read
  joins by id without an active filter); they just leave the RATED set and
  the default picker.
- Trim the default BINARY tag set roughly in half: deactivate the
  sleep-tag category (now covered by the sleep-quality slider), the mood
  level restatements, the nutrition long-tail, and low-signal day-type
  tags. Kept rows are the high-signal, correlatable tags. Nothing is
  deleted; inactive rows still resolve labels for existing entries.

Reference-data only — no schema change, idempotent upsert + closed-key
updates, no per-user data touched.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

messages/de.json

@@ -577,7 +577,8 @@
       "factorSleepQuality": "Schlafqualität",
       "factorStress": "Stress",
       "factorConflict": "Konflikt",
-      "factorFamily": "Familie"
+      "factorFamily": "Familie",
+      "factorSadness": "Traurigkeit"
     }
   },
   "medications": {

messages/en.json

@@ -577,7 +577,8 @@
       "factorSleepQuality": "sleep quality",
       "factorStress": "stress",
       "factorConflict": "conflict",
-      "factorFamily": "family"
+      "factorFamily": "family",
+      "factorSadness": "sadness"
     }
   },
   "medications": {

messages/es.json

@@ -577,7 +577,8 @@
       "factorSleepQuality": "calidad del sueño",
       "factorStress": "estrés",
       "factorConflict": "conflicto",
-      "factorFamily": "familia"
+      "factorFamily": "familia",
+      "factorSadness": "tristeza"
     }
   },
   "medications": {

messages/fr.json

@@ -577,7 +577,8 @@
       "factorSleepQuality": "qualité du sommeil",
       "factorStress": "stress",
       "factorConflict": "conflit",
-      "factorFamily": "famille"
+      "factorFamily": "famille",
+      "factorSadness": "tristesse"
     }
   },
   "medications": {

messages/it.json

@@ -577,7 +577,8 @@
       "factorSleepQuality": "qualità del sonno",
       "factorStress": "stress",
       "factorConflict": "conflitto",
-      "factorFamily": "famiglia"
+      "factorFamily": "famiglia",
+      "factorSadness": "tristezza"
     }
   },
   "medications": {

messages/pl.json

@@ -577,7 +577,8 @@
       "factorSleepQuality": "jakość snu",
       "factorStress": "stres",
       "factorConflict": "konflikt",
-      "factorFamily": "rodzina"
+      "factorFamily": "rodzina",
+      "factorSadness": "smutek"
     }
   },
   "medications": {

prisma/migrations/0126_v1129_mood_catalog_revision/migration.sql

@@ -0,0 +1,72 @@
+-- v1.12.9 — mood-catalog revision.
+--
+-- Tightens the operator-finalised mood factor + tag catalog seeded by
+-- 0118 / 0119 / 0120. Three changes, all reference-data only — no schema
+-- change, no per-user data touched, no destructive delete:
+--
+--   1. Seed a new RATED factor `factor_sadness` (scale 1..5, inverse —
+--      higher = a worse day), under the existing `factors` category.
+--   2. Retire the four RATED factors the operator dropped
+--      (`factor_conflict`, `factor_family`, `factor_social`,
+--      `factor_stress`): flip each to `kind = 'BINARY'` AND set
+--      `is_active = false`. They leave the RATED set and the default
+--      picker, but the catalog rows stay so any historical
+--      `mood_entry_tag_links` row still resolves its label (the insights
+--      read joins the catalog row by id and does NOT filter `is_active`).
+--      The result: exactly THREE RATED factors remain
+--      (`factor_work`, `factor_sleep_quality`, `factor_sadness`).
+--   3. Trim the default BINARY tag set roughly in half: mark the
+--      long-tail / slider-redundant tags `is_active = false`. The picker
+--      (`GET /api/mood/tags` filters `is_active = true`) stops offering
+--      them; existing entries that reference a now-inactive tag still
+--      resolve its label because the read joins by id without the active
+--      filter. Nothing is deleted.
+--
+-- Idempotent: the seed uses a deterministic id + ON CONFLICT upsert by
+-- `key` (mirrors 0118 / 0119 / 0120); the retire / trim updates key off a
+-- closed `key` list, so a re-run is a no-op.
+
+-- ─── 1. New RATED factor: sadness (inverse, 1..5) ────────────────────
+INSERT INTO "mood_tags"
+  ("id", "category_id", "key", "label_key", "icon", "sort_order", "is_active", "kind", "scale_min", "scale_max", "inverse")
+VALUES ('mt_factor_sadness', 'mtc_factors', 'factor_sadness', 'mood.tag.factorSadness', 'CloudRain', 2, true, 'RATED', 1, 5, true)
+ON CONFLICT ("key") DO UPDATE SET
+  "category_id" = EXCLUDED."category_id", "label_key" = EXCLUDED."label_key", "icon" = EXCLUDED."icon",
+  "sort_order" = EXCLUDED."sort_order", "is_active" = EXCLUDED."is_active", "kind" = EXCLUDED."kind",
+  "scale_min" = EXCLUDED."scale_min", "scale_max" = EXCLUDED."scale_max", "inverse" = EXCLUDED."inverse";
+
+-- Re-seat the two surviving RATED factors at the top of the `factors`
+-- category so the final order reads work → sleep quality → sadness.
+UPDATE "mood_tags" SET "sort_order" = 0 WHERE "key" = 'factor_work';
+UPDATE "mood_tags" SET "sort_order" = 1 WHERE "key" = 'factor_sleep_quality';
+
+-- ─── 2. Retire the four dropped RATED factors (BINARY + inactive) ─────
+-- Flip kind so they leave the RATED set; deactivate so they leave the
+-- default picker. Historical RATED links keep their `rating` value and
+-- still resolve their label (read joins the catalog row by id, no active
+-- filter). Not deleted — references stay valid.
+UPDATE "mood_tags"
+SET "kind" = 'BINARY', "is_active" = false
+WHERE "key" IN ('factor_conflict', 'factor_family', 'factor_social', 'factor_stress');
+
+-- ─── 3. Trim the default BINARY tag set (~half) ──────────────────────
+-- Deactivate the long-tail / slider-redundant binary tags. Kept set is
+-- the high-signal, plausibly-correlatable tags; dropped set is sleep
+-- tags now covered by the Schlafqualität slider, mood-level restatements
+-- (the 1..5 face already carries them), and low-signal duplicates. Rows
+-- stay so historical entries still resolve their labels.
+UPDATE "mood_tags"
+SET "is_active" = false
+WHERE "key" IN (
+  -- feelings: drop tags that just restate the 5-face mood level / overlap
+  -- the surviving factor_sadness; keep a tight high-signal subset.
+  'content', 'unsure', 'bored', 'tense', 'stressed', 'sad',
+  -- sleep: the Schlafqualität (factor_sleep_quality) slider supersedes
+  -- the whole sleep-tag category.
+  'slept_well', 'slept_ok', 'slept_poorly', 'early_night',
+  -- health: drop the nutrition long-tail (kept: ate_well, alcohol,
+  -- worked_out, walked, hydrated).
+  'fast_food', 'no_sweets', 'big_meal',
+  -- work: drop low-signal day-type tags (kept: productive, day_off).
+  'overtime', 'travel', 'sick_day'
+);

src/components/mood/mood-tag-icons.ts

@@ -10,6 +10,7 @@ import {
   CheckCircle,
   Clock,
   CloudMoon,
+  CloudRain,
   CloudSun,
   Dumbbell,
   Film,
@@ -66,6 +67,7 @@ const MOOD_TAG_ICONS: Record<string, LucideIcon> = {
   CheckCircle,
   Clock,
   CloudMoon,
+  CloudRain,
   CloudSun,
   Dumbbell,
   Film,