Meals & Shopping
Foundation in progress
A web application for household meal planning: ingest recipes, collect family votes, choose winners, aggregate ingredients by store section, and present cook-ready meal instructions.
Product
Meals & Shopping is a practical household app: find candidate meals, get input from the family, combine the winning recipes into one shopping list, and make the cooking step less chaotic.
The product idea is explicitly AI-shaped. The system should learn preferences over time, propose meals that fit the household, extract structured ingredients from messy sources, and flag low-confidence recipes instead of silently creating a bad list.
The current repo is laying the foundation: FastAPI, Vite/React, PostgreSQL, Docker deployment, household tenancy, invite links, and Playwright-driven verification. Later phases add the extraction, voting, shopping, pantry, and cook-mode engines.
What it shows
- Family members vote on proposed meals instead of one person carrying the whole decision.
- Recipe links, videos, photos, and follow-on pages are planned as ingestion sources.
- Winning meals roll into an aisle-grouped grocery list with quantity aggregation.
- Cook mode is scoped to surface ingredients, prep warnings, timers, video, and steps when it matters.
Stack
- FastAPI
- React
- Vite
- PostgreSQL
- Docker
- Playwright
- Recipe extraction
- Agent plans