Real-time shared lists

• Firestore-backed, grouped by store with per-store subtotals and live savings. Add, check off, adjust quantity; every change syncs to all members in the household. Daily offers from 5 chains
• a GitHub Actions cron scrapes each retailer every morning at 07:00 (VTEX/GraphQL APIs where available, HTML fixtures where not) and publishes a normalized offer feed. Full-catalog live search
• for Auchan and Mega Image (which expose their whole stock), search hits their APIs live for ~60k in-stock products, not just the weekly deals. Cross-store price comparison
• tap any offer to see the same product at other chains, matched on name tokens and unit, cheapest first. Unit prices & smart filters
• price-per-kg/l parsed from titles, filter by store / category / price range / discount state, sort by price, discount, or unit price. Multiple households
• create and switch between groups; owners can rename the group and manage members; invites are a single shareable link. Attribution
• see who added each item and who checked it off, by avatar. Shopping sessions & history
• finishing a shop archives the checked items into a dated session with the total spent and who took part. Installable PWA
• custom Firebase auth handler served from the app's own domain, offline-capable, home-screen install on iOS/Android. Built with

Cost-aware data layer

• offers are read from the persistent cache and refreshed at most once per day per user; the scraper diffs against existing docs and only writes what changed, keeping Firestore well within the free quota despite thousands of offers. Race-safe cart
• adding the same offer repeatedly increments a single deterministic document inside a Firestore transaction, so rapid taps become a quantity, not duplicate rows. Honest scraping
• each retailer was reverse-engineered from its own frontend; where a chain exposes only its weekly leaflet (Penny/Kaufland/Lidl) the app is explicit about it, and only the two chains with real catalog APIs power full-stock search.