feat(profile): user-profile feature registry + builder (phase A)

Centralizes user-level features (completion_rate_30d, dismiss_rate_30d,
mean_dwell_ms_30d, preferred_hour, tip_volume_30d) in a TS registry that
owns both definition and SQL aggregation, since the data lives in the
TS-owned SQLite tables (tip_views/tip_feedback). Lazy TTL refresh keeps
recommend latency bounded; values persist in user_profile_features (KV).

ml/serving accepts profile_features on /score + /generate but does not
yet consume them — extending the bandit feature vector changes D and
resets every user's learned state, so that's a deliberate phase-B step.

Includes ml/features/profile_schema.py as a contract mirror with a sync
test that diffs name sets against registry.ts.

ADR-0011 records the data-locality reasoning (registry in TS, not Python
as the issue originally suggested).

Phase B (deferred): event-driven incremental updates, bandit consumption
with state migration, admin per-user profile page, staleness alerts.

Refs #81.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

services/api/src/db/schema.ts

@@ -87,6 +87,20 @@ export const tipScores = sqliteTable('tip_scores', {
   tipKind: text('tip_kind'),                         // 'task' | 'advice' | 'insight' | 'reminder'
 });
 
+// ── User profile features (#81 phase A) ────────────────────────────────────
+// One row per (userId, name). KV store for aggregated user-level features
+// computed from tip_views/tip_feedback/tip_scores. Numeric values land in
+// `value`; categorical/string values use `value_text` (never both). Entries
+// are recomputed lazily by the profile builder when older than `ttl_sec`.
+export const userProfileFeatures = sqliteTable('user_profile_features', {
+  userId: text('user_id').notNull().references(() => users.id),
+  name: text('name').notNull(),                       // e.g. 'completion_rate_30d'
+  value: integer('value', { mode: 'number' }),        // numeric (REAL stored as number); null if categorical
+  valueText: text('value_text'),                       // categorical/string; null if numeric
+  updatedAt: text('updated_at').notNull(),
+  ttlSec: integer('ttl_sec').notNull(),                // staleness threshold; 0 = never auto-refresh
+});
+
 // ── Simulation runs ──────────────────────────────────────────────────────────
 // One row per offline simulation run (two-policy comparison).
 export const simRuns = sqliteTable('sim_runs', {