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/Vision.mdBedrockactive2026-07-16Read when: at every phase open (Opening Checklist Ring 1); before any major decision

Vision

The Thesis

People do not kill houseplants through neglect. They kill them by following care advice written for someone else's room.

"Water weekly" is wrong in a dark hallway and wrong again by a south window. Frond measures the room, then schedules the plant. The claim is that a schedule you can see the reasoning for is a schedule you will actually follow.

Who it's for

Someone with between three and fifteen houseplants who has already killed a few and minds about it. Not a collector with a humidity cabinet, and not someone who wants a single rubber plant to survive by luck.

The job they are hiring us for: "tell me what to do this week, and be right often enough that I stop second-guessing you."

What success looks like

A user who has not opened the app in a month still waters on our schedule, because they learned it.

The skeptic-convincer is narrower: a user who moves a plant to a new room, re-surveys, sees the schedule change, and agrees with the change.

Principles

  • Never guess a schedule. One honest question beats an invented number.
  • Copy never scolds. A missed plant is a failure of our timing, not the user's character.
  • Offline first. The app has to work where the plants are.
  • Confidence is a feature. Show why a schedule says what it says, or it will not be trusted.