diff --git a/prisma/migrations/0073_v1439_medication_intake_unique_per_source/migration.sql b/prisma/migrations/0073_v1439_medication_intake_unique_per_source/migration.sql
new file mode 100644
index 000000000..df174c3f3
--- /dev/null
+++ b/prisma/migrations/0073_v1439_medication_intake_unique_per_source/migration.sql
@@ -0,0 +1,57 @@
+-- v1.4.39 W-SERVER-FIX-2 — unique constraint on
+-- (user_id, medication_id, scheduled_for, source) for
+-- medication_intake_events so the projection-then-backfill code path
+-- can't mint duplicate REMINDER rows under a concurrent two-route
+-- hit.
+--
+-- Background. Two read routes now project active schedules through
+-- `expandTodayIntakes` and idempotently `createMany`-backfill any
+-- missing rows:
+--
+--   GET /api/medications/intake?scope=today
+--   GET /api/dashboard/summary
+--
+-- The iOS Dashboard mounts both at roughly the same wall-clock when
+-- the compliance ring + Erfassen sheet hydrate together. Pre-fix the
+-- table carried only an INDEX on `(user_id, medication_id,
+-- scheduled_for)`, NOT a UNIQUE constraint, so both routes could pass
+-- their existence probes simultaneously and `createMany` the same
+-- `(medicationId, scheduledFor, REMINDER)` row twice — corrupting the
+-- compliance bucket counts on the next read.
+--
+-- `source` is part of the key so the reminder worker's REMINDER row
+-- doesn't collide with manual user inserts (WEB / TELEGRAM / IMPORT)
+-- on the same slot. Distinct intake sources at the same scheduled
+-- instant are a legitimate shape — e.g. a Telegram-bot logged take +
+-- a Web-UI correction on the same dose.
+--
+-- The query-side `(user_id, medication_id, scheduled_for)` index
+-- stays in place — most reads filter on those three columns without
+-- caring about `source`, so the legacy index keeps its read-path
+-- value even after the unique constraint creates its own
+-- (user_id, medication_id, scheduled_for, source) covering index.
+--
+-- Production data check. The reminder worker is the only writer that
+-- ever produces REMINDER-source rows pre-this-fix; it queries pending
+-- rows by `(medicationId, scheduledFor)` before each insert so a
+-- self-double-mint is structurally impossible. The new projection
+-- code paths (intake route + dashboard route) are the only new
+-- sources of potential duplicates, and both ship with the
+-- `skipDuplicates: true` flag alongside this constraint. If
+-- `migrate deploy` reports a unique-constraint violation, a follow-up
+-- one-shot data-cleanup pass would dedup the offending rows by
+-- (user_id, medication_id, scheduled_for, source) keeping the
+-- lowest-id row; that cleanup is intentionally NOT bundled here so
+-- the migration stays additive + reviewable.
+--
+-- Additive only. IF NOT EXISTS / EXCEPTION-on-duplicate guards
+-- mirror the patterns in 0067 / 0071.
+
+DO $$ BEGIN
+    ALTER TABLE "medication_intake_events"
+        ADD CONSTRAINT "medication_intake_events_user_id_medication_id_scheduled_for_source_key"
+        UNIQUE ("user_id", "medication_id", "scheduled_for", "source");
+EXCEPTION
+    WHEN duplicate_object THEN NULL;
+    WHEN duplicate_table THEN NULL;
+END $$;
diff --git a/prisma/schema.prisma b/prisma/schema.prisma
index 328e04373..b0509aee8 100644
--- a/prisma/schema.prisma
+++ b/prisma/schema.prisma
@@ -838,6 +838,15 @@ model MedicationIntakeEvent {
   user       User       @relation(fields: [userId], references: [id], onDelete: Cascade)
   medication Medication @relation(fields: [medicationId], references: [id], onDelete: Cascade)
 
+  /// v1.4.39 W-SERVER-FIX-2 — both the intake route and the dashboard
+  /// summary route project active schedules through `expandTodayIntakes`
+  /// and idempotently `createMany`-backfill missing rows. Without a
+  /// unique constraint the two reads could race a duplicate REMINDER
+  /// row in for the same `(medicationId, scheduledFor)` between the
+  /// existence probe and the insert. `source` is in the key so the
+  /// reminder worker (REMINDER) doesn't collide with manual user
+  /// inserts (WEB / TELEGRAM / IMPORT) on the same slot.
+  @@unique([userId, medicationId, scheduledFor, source])
   @@index([userId, medicationId, scheduledFor])
   @@map("medication_intake_events")
 }