Connect Nutrition Research to ChatGPT

Connects the nutrition literature to the products people actually eat.

Tools:

• search_nutrition_studies: PubMed narrowed with MeSH terms (Nutritional Sciences, Diet, Dietary Supplements, Nutrients), so a broad word like 'protein' returns nutrition science rather than unrelated biochemistry. Metadata and links only.
• search_nutrition_reviews: Europe PMC filtered to systematic reviews and meta-analyses. Dietary questions are better answered by pooled evidence than by single trials, so reviews are the default.
• search_food_products: Open Food Facts, restricted to products sold in Japan, returning declared nutrition per 100 g (energy, protein, fat, carbohydrate, sugars, fiber, salt, sodium).
• get_food_product_by_barcode: the same data for one product by barcode; an unregistered barcode returns product=null rather than an error.

Data caveats: Open Food Facts values are transcribed from product labels by contributors, so treat them as label data, not laboratory measurements, and expect gaps. Abstracts are returned only for open-access articles under a CC license; other records carry metadata and a link.

This server surfaces published findings and label values. It does not assess a diet, diagnose deficiency, or give medical or dietetic advice — anyone with a medical condition or specific dietary need should consult a physician or registered dietitian.