The model cannot write directly
It proposes bounded actions. The API validates ownership, date scope, domain rules, and values before anything is stored.
Safety & transparency
Helpful enough to use. Honest enough to question.
AI is part of the interaction, but it is not the authority over your account, calculations, or wellbeing.
Estimates stay inspectable
Food values are approximate and visibly editable. Unknown micronutrients remain unknown instead of becoming fabricated zeros.
Account controls remain server-side
Authentication, authorization, session checks, totals, targets, deficits, and report snapshots are controlled by trusted backend code.
Ambiguity gets a focused question
When identity or quantity would make an estimate irresponsible, NectariaFit asks for the missing detail instead of guessing silently.
Sensitive requests are screened
Explicit medical disclosures, crisis content, secrets, and narrowly defined sensitive data are screened before full account context is loaded.
Wellness has a clear boundary
NectariaFit is not a healthcare provider, medical device, diagnostic service, treatment program, or emergency channel.
Supported
Everyday wellness context
- Food, portions, calories, and macronutrients
- Ordinary activity, distance, and duration
- Weight, height, age, routine activity, and goals
- Corrections, deletions, and progress questions
Not supported
Medical and highly sensitive use
- Diagnoses, symptoms, medication, or test interpretation
- Pregnancy, injury, allergy, or eating-disorder information
- Crisis, self-harm, dangerous restriction, or emergencies
- Passwords, payment data, identity documents, or another person’s private data
Read the full details.
The public documents explain data handling, AI limitations, account responsibilities, and the legal boundaries of the service.