Meal Suggestions combine remaining macros, timing, and schedule reality so the next meal is practical instead of theoretically perfect. The job is not to produce endless options. It is to narrow the field to choices you are actually likely to eat and log.
01Recommendation score
| Source | Score basis | Typical advantage |
|---|
| Nutrition quality | protein, fiber, and micronutrient spread | stronger consistency and satiety |
| Convenience | prep time and ingredient access | better adherence under pressure |
| Adherence risk | prep complexity and repeat fatigue | reduces likelihood of fallback to poor options |
02Accept and override examples
| Context | Recommended action |
|---|
| Good signal match | accept primary suggestion and log one full serving |
| Time squeeze with low quality stock | use a simpler convenience fallback and keep protein target |
| Repeated mismatch with appetite | override with higher satiety profile and log reason |
03Low-quality output correction
| Failure pattern | Trigger count | Correction |
|---|
| Repeated low-satiety picks | 3 consecutive recommendations | raise food quality weight in next cycle |
| Multiple missed follow-up logs | 2 weeks | shift to stricter simple templates |
| High sodium or sugar bias | repeated and unaligned | cap that class and force alternative categories |
04What should shape the next suggestion
| Input type | Why it matters |
|---|
| Remaining protein | protects recovery and fullness late in the day |
| Remaining calories | prevents one meal from forcing panic adjustments |
| Time available | determines whether prep-heavy advice will fail |
| Inventory and access | keeps suggestions grounded in the food you can get |
| Appetite state | changes whether volume or density should lead |
Anchor suggestions with macro tracking, macro-friendly recipes, and meal planning for reliability.
