Point at a product and see if you already own it

Build point at a product and see if you already own it.

Context: You're in a shop holding a charging cable, or a screwdriver set, or a spice you
half remember buying. Point your phone at it and get one answer: you already
have one, at home, bought in March.

Everyone has bought the same thing twice. Cables, batteries, the identical
white t-shirt, a book you already read. It isn't forgetfulness, it's that
nobody has an inventory of their own flat and there's no realistic way to build
one by hand.

So it can't ask people to catalogue anything. That's the whole design problem.
It has to build the inventory from things they already do — receipt emails,
order confirmations, delivery notifications. Those arrive automatically and
almost nobody deletes them.

The recognition side is the easier half: matching what you're pointing at
against a list of things you've bought is a lot simpler than identifying an
object from nothing.

The hard part is trust. It only works if it reads your email, and that is a lot
to ask. It probably has to be local-first, or scoped so tightly that it only
ever touches order confirmations and can prove it.

Second hard part: knowing whether you still have it. Bought in 2019 doesn't mean
it's in the drawer. Better to say "you bought one in March 2024" and let the
person decide than to claim more than it knows.

Requirements:
- Ship the smallest version that delivers the core value above, nothing more.
- Pick a stack you can deploy in a day, and state the choice before writing code.
- Handle the empty state, the error state and the offline case explicitly.
- Pricing to build toward: Free for the basics, since it needs to be there in the aisle before people
trust it. $5/mo for the full history and household sharing — two people buying
the same thing is the more common version of this problem anyway.

Start by listing the screens, then build them one at a time.