Example project — Frond is not real.

Authored demo content for a fictional plant-care app, rendered by the same parsers and held to the same invariants as a real project's dashboard. This is the system in use, three months in.

docs/strategy/Scope & Constraints.mdCommitmentsactive2026-07-16Read when: at phase open when scope is in question; before queuing a new phase

Scope & Constraints

In scope (first milestone)

A single user tracking houseplants in one home: add a plant, survey the room it lives in, receive a daily digest of what needs water, and mark it done.

Out of scope (deliberate non-goals)

  • Outdoor gardening. Weather, seasons and soil turn it into a different product.
  • Plant identification from a photo. The care model has to earn trust before we point a camera at a stranger's plant.
  • A social feed. Plant people are lovely and a feed is a different company.
  • Fertiliser, repotting and pest schedules. Water is the habit; everything else rides on it later.
  • Multi-user households. Queued as a direction, not a first milestone.

Hard constraints

  • Two people, part time. Scope that needs a botanist or a data curator is out by definition.
  • Species data is licensed, so we cannot correct a bad entry directly.
  • Push notifications are a finite budget. One digest a day is the ceiling we set ourselves.
  • Northern hemisphere only in the current seasonal logic. Known, embarrassing, tracked.

Assumptions

  • (Riskiest) A phone camera can distinguish light bands well enough to change a schedule. If it cannot, the whole "measure the room" premise collapses into a questionnaire. The sensor accuracy sweep is queued to test it.
  • Users will complete a room survey once per plant if it takes under a minute.
  • A daily digest is enough urgency for a plant that is two days overdue.