Initial commit: full fitness-rewards app

Backend: FastAPI + PostgreSQL + SQLAlchemy async
- Auth (JWT), onboarding (PAR-Q+, TTM, BREQ-2), activities, food log
- Points engine with daily gate + weekly targets
- Strava + Polar OAuth integration
- Open Food Facts barcode lookup
- Resource recommendation engine
- 10 seeded Darebee/StrongLifts/YouTube programs

Frontend: React + Vite, mobile-first dark theme
- Login/Register, 8-step onboarding flow
- Dashboard with daily gate status
- Activity logger, food logger (manual + barcode scan + search)
- Reward shop with redemption
- Weekly summary view
- Settings (point rules, targets, integrations)

Deployment: Docker Compose for Coolify (Traefik)

backend/app/services/resource_matcher.py

@@ -0,0 +1,67 @@
+"""Resource recommendation engine — matches external programs to user profiles."""
+import uuid
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+from app.models.resource import Resource
+from app.models.profile import UserProfile
+
+
+async def get_recommendations(db: AsyncSession, user_id: uuid.UUID) -> list[dict]:
+    profile_result = await db.execute(
+        select(UserProfile).where(UserProfile.user_id == user_id)
+    )
+    profile = profile_result.scalar_one_or_none()
+    if not profile:
+        return []
+
+    query = select(Resource).where(Resource.is_active == True)
+    if profile.equipment_access == "none":
+        query = query.where(Resource.equipment_needed == "none")
+    elif profile.equipment_access == "basic_home":
+        query = query.where(Resource.equipment_needed.in_(["none", "basic_home"]))
+
+    result = await db.execute(query)
+    resources = result.scalars().all()
+
+    user_prefs = set(profile.activity_preferences or [])
+    user_goal = profile.primary_goal
+    user_stage = profile.ttm_stage
+
+    difficulty_map = {
+        "precontemplation": "beginner",
+        "contemplation": "beginner",
+        "preparation": "beginner",
+        "action": "intermediate",
+        "maintenance": "advanced",
+    }
+
+    scored = []
+    for r in resources:
+        score = 0
+        if user_goal and user_goal in (r.goal_tags or []):
+            score += 30
+        overlap = user_prefs & set(r.activity_tags or [])
+        score += len(overlap) * 15
+        if user_stage and user_stage in (r.ttm_stages or []):
+            score += 20
+        if r.difficulty == difficulty_map.get(user_stage):
+            score += 10
+        scored.append((r, score))
+
+    scored.sort(key=lambda x: x[1], reverse=True)
+
+    return [
+        {
+            "id": str(r.id),
+            "name": r.name,
+            "source": r.source,
+            "url": r.url,
+            "description": r.description,
+            "difficulty": r.difficulty,
+            "duration_days": r.duration_days,
+            "goal_tags": r.goal_tags,
+            "activity_tags": r.activity_tags,
+            "score": s,
+        }
+        for r, s in scored[:8]
+    ]