All LLM parameters must now be explicitly set in .env — no silent
fallbacks in code. Added nvidia_enable_thinking bool and wired it into
the chat completion extra_body instead of the previous hardcoded True.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Web app React 18 + TypeScript con quattro schermate: Setup (URL server +
password + test connessione), Ingredients (lista ingredienti e preferenze),
Recipes (lista 3 ricette generate), RecipeDetail (dettaglio e procedimento).
Persistenza URL/password via Capacitor Preferences. Pronto per build Android.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Python 3.11 + FastAPI con tre endpoint (/health, /auth/test, /recipes/generate).
Autenticazione via header X-App-Password, validazione input Pydantic,
chiamata in streaming a NVIDIA NIM (nemotron-3-ultra-550b-a55b) con thinking.
Dockerizzato con docker-compose; configurazione tramite .env.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
davide