Transparency
Procaaf is a free scratchpad for composing plates and seeing protein, carbs, fats and calories update live. This page explains exactly how those numbers are produced, how accurate they are, and where the limits of that accuracy lie.
Procaaf is an educational calculator. It is not a medical device, and nothing on this site is medical, clinical, or dietetic advice, diagnosis, or treatment. Individual needs vary with age, health conditions, medication, pregnancy, training load and more. Talk to a qualified doctor or registered dietitian before changing your diet, especially if you have diabetes, kidney or liver disease, an eating disorder, or any other medical condition. Never delay or disregard professional advice because of something you read here.
Every food entry stores protein, carbohydrate and fat in grams per 100 g (or per 100 ml for liquids). Values are compiled from publicly available food composition references — chiefly the USDA FoodData Central database, supplemented by regional composition tables and typical manufacturer label values for packaged and regional items that have no generic reference entry.
Entries are stored as generic, representative foods (for example “chicken breast, cooked”), not specific brands. Where sources disagree, we use a mid-range value rather than the highest or lowest.
For each item on your plate we scale the reference values linearly by the portion you enter:
grams of macro = (macro per 100 g ÷ 100) × your portion in grams
Plate totals are the sum of every item. Percentages shown in the ring and donut are each macro's share of total calories, not of total weight.
Calories are derived from macros using the standard Atwater factors, not stored separately:
kcal = (protein g × 4) + (carbs g × 4) + (fat g × 9)
Because of this, Procaaf's calorie figure can differ by a few percent from a product label. Labels may use specific Atwater factors, count fibre or sugar alcohols differently, or round each value independently. Alcohol (7 kcal/g) is not modelled.
Pairing suggestions are based on the classic limiting–amino-acid model: most plant proteins are relatively low in either lysine (grains, nuts, seeds) or methionine (legumes), so combining the two groups across a day gives a more complete amino acid profile. Suggestions are qualitative guidance for variety — they are not a per-milligram amino acid computation, and no scoring method (PDCAAS/DIAAS) is applied.
Daily macro targets you set are simply reference lines: the app compares your plate against the number you typed and does not generate a prescription for you. The BMI calculator uses the standard formula (weight in kg ÷ height in metres squared) with WHO adult categories. BMI ignores body composition, muscle mass, frame, ethnicity, age and sex, so it is a rough screening figure only — it is not a diagnosis and not a body-fat measurement.
Cooking method, moisture loss, trimming, ripeness, cut, breed, soil and brand all move real macro values. Treat every number here as a good estimate, typically within roughly 10% of the true value for a generic food, and further off for mixed or restaurant dishes. Micronutrients, fibre, sodium and sugars are not tracked. Portions you enter are cooked-or-raw as you interpret them; the app cannot tell the difference.
Food data is reviewed as the database grows. If a value looks wrong, email procaaf@gmail.com with the food and your source, and we'll check it.