Procedure

1. Baseline Assessment

   • Call get_user_preferences and get_user_history to identify current body composition goals (fat loss, maintenance, or hypertrophy).
   • If no baseline exists, calculate expected maintenance energy using compute_tdee.
   • Inform the user that the first phase involves establishing a caloric baseline to account for metabolic variance and non-exercise activity thermogenesis (NEAT).

2. Macronutrient Prioritization (The Anchor)

   • Protein: Set protein as the primary target. For muscle gain, target 1.6–2.2g/kg of body weight. For fat loss, ensure high protein intake to retain lean mass.
   • Fats/Carbs: Allocate 15–30% of total calories to fats. Assign the remaining calories to carbohydrates to support training performance.
   • Use set_user_macro_targets to save these parameters.

3. Meal Plan Construction

   • Search for appropriate recipes or ingredients using search_foods and find_meal_candidates.
   • Prioritize whole foods for ~80% of caloric intake, leaving ~20% for flexible choices to improve long-term adherence.
   • Implement create_meal to build a multi-week schedule focused on weekly averages rather than daily perfection.

4. Flexible Substitution Logic If a user requests an ingredient swap or an item is unavailable, follow this hierarchy:

   • Macronutrient Match: Suggest items with similar protein/calorie density (e.g., swapping chicken breast for tilapia).
   • Volume/Satiety Swaps: For fat loss, suggest low-density swaps like cauliflower rice for white rice or spaghetti squash for pasta.
   • Functional Compatibility: Suggest swaps that preserve texture, such as Greek yogurt instead of sour cream.
   • Use generate_recipe to adjust instructions based on the new ingredient.

5. Adaptive Feedback Loop (2–4 Week Cycle)

   • Use get_week_stats every 2–4 weeks to monitor average weight and performance markers.
   • Compare progress against the user's primary goal.
   • If progress plateaus, update the plan by adjusting caloric intake or activity levels to restart progress.
   • Avoid labeling foods as "off-limits" to prevent the psychological pitfalls of rigid dieting.

6. Review and Log

   • Render the current plan using render_meal_log to show the user their upcoming week.
   • Encourage the user to focus on weekly macro averages to accommodate social events or refeed days.