July 30, 2026 · Handbook

Manual: the Discovery page

The earlier chapter covered how Opportunity Briefs get made. This one is the desk manual: what to do with the page they land on.

Reading a brief skeptically

A brief's anatomy — problem, audience, evidence, MVP scope, confidence — is designed for skeptical reading, in this order:

  • Evidence first. Real quotes, real threads, real numbers? The agents must show their sources; a brief is only as good as what it links.
  • Audience second. "Who exactly, and where do they gather" — if that's vague, the marketing half of the loop will have nothing to aim at later.
  • Scope last. A good MVP scope is uncomfortably small. If it reads like a quarter's roadmap, trim it in the build prompt — you're allowed to build less than the brief suggests.

The three things you can do with one

ActionWhen
Build thisConvinced. The brief becomes the build prompt — plan, approve, done; the normal build flow from there.
Validate firstHalf-convinced. Stage a waitlist page and a paused ad campaign to measure pull before building — a dud idea costs a landing page.
Let it ageNot convinced. Briefs keep; the queue is a reading list, not a to-do list, and ignoring most of it is the correct usage.

Keeping the queue worth reading

Discovery runs on demand or on a schedule — weekly suits most accounts. Steer it by telling it your territories: research runs accept a topic focus, so "meal planning" or "tools for climbing gyms" sweeps narrow instead of boiling the internet. The narrower the territory, the sharper the briefs.

The habit that compounds: when you reject a brief, notice why — wrong audience, weak evidence, too big — and put that word in your next research run's focus. The queue gets better at being yours.
← All posts