agentleFS
mannobeatsmannobeats/hyper-taste/skills/product-brief/SKILL.md

product-brief

Skill1 starsChanged 4 days agoMITView on GitHub

One-page brief (docs/BRIEF.md) for a new app or product, written before any code - what it is, who it is for, the main jobs, the data, the platform and stack, the look, what is out of scope, and how we will know it works. Use when the user has an idea, wants to replace an app they use, hands over a rough spec, or research is done.

12 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use decision-log and setup.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • nandanosqlSkill

    decision-log

    Document product decisions with context, alternatives considered, rationale, and revisit criteria. Creates institutional memory that prevents re-litigating settled decisions.

    Also in 3 projects that use this

  • VercelSkill

    setup

    Set up Vercel CLI and project configuration. Use when the user says "set up Vercel", "configure Vercel", "link to Vercel", or "vercel init".

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 2 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 8 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Write docs/BRIEF.md in a folder named after the product (<parent>/<Name>/docs/BRIEF.md.

    line 8
  2. Very specific:

    If docs/research/ or docs/screens/INDEX.md exist, read them first.

    line 17
  3. Very specific:

    If the user gives you a spec or PRD, it already answers most questions: copy what fits, keep the rest as docs/SPEC.md, and ask only what is …

    line 20
  4. Distinctive:

    Ask one or a few questions at a time, each with your best guess, so "I don't know" still moves things forward.

    line 25
  5. Distinctive:

    Should it look like the platform's own apps, or have its own look?

    line 41
  6. Very specific:

    Call the Skill tool with user-stories, For a product that agents will build over many sessions, also call it with quality-bar.

    line 93

“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Product brief

Write docs/BRIEF.md in a folder named after the product (<parent>/<Name>/docs/BRIEF.md; ask for the parent folder, default ~/Developer). Every later skill reads it. Keep it to one page. It says what to build and why, before anyone argues about how.

1. Name and read what exists

Pick the name first: one word (Tasks, DailyLog), not the name of an app the person already uses, and for an app a reverse-DNS id (com.<yourname>.<appname>) in lowercase. The folder, the research, and the project all use it.

If docs/research/ or docs/screens/INDEX.md exist, read them first. Skip every question they already answer.

If the user gives you a spec or PRD, it already answers most questions: copy what fits, keep the rest as docs/SPEC.md, and ask only what is missing.

2. Interview

Ask one or a few questions at a time, each with your best guess, so "I don't know" still moves things forward. Never ask what the code or the research already says. Ask what the person actually needs, not only what they asked for: "a dashboard" is often "know which clients pay late". If the user asked you to decide everything, answer each question yourself from the research, write the choice in the brief, and note "decided by the agent" next to it.

  1. What do you do today, and which app or habit does it replace?
  2. What are the three things you do most often? These become the main pages.
  3. What data does it hold? Name the nouns: tasks, notes, recipes, invoices.
  4. Is there a reference app you like? Then the clone skill comes next.
  5. Where does it live: a Mac app (a window with a sidebar, a menu bar item, a quick-capture panel), a web app, a phone app, or a mix? Which stack (SwiftUI, Tauri, Electron, a web framework, Expo, native mobile)?
  6. Does it need the internet? Default: no.
  7. Which three things are out of scope for version one?
  8. Should it look like the platform's own apps, or have its own look? Note a brand or accent color, and any app they name ("like Things"). No answer means native, with the system's accent color.
  9. How will we know it works? One or two outcomes a person could check.

3. Cut scope

A personal app has one user. Remove what exists only for other users: accounts, sharing, teams, sync servers, billing, onboarding, analytics. A product for many users keeps what they need, and nothing for the vendor's sake.

Version one is small enough to finish: two to five main pages, one data model with two to four types. Smart lists that filter one model (Today, Upcoming, Someday) count as one page.

4. Write docs/BRIEF.md

# <AppName>

One sentence: what it is and who it is for.

## Replaces
<app or habit>

## Main jobs
1. ...
2. ...
3. ...

## Data
- <Type>: <fields>

## Shape
<sidebar window / menu bar extra / web app / phone app>, <platform and stack, and why>. <Data: local or synced. Offline or not. Accounts or not.>

## System hooks
<global shortcut, menu bar, Share menu, Shortcuts, CLI, or "none">

## Reference
<app name, or "none">

## Look
<native with an accent color (name it, or "system"), its own look, or "like <app>", and why>

## Works when
- <an outcome a person can check>

## Not in version one
- ...

5. Hand off

Call the Skill tool with user-stories, For a product that agents will build over many sessions, also call it with quality-bar.

Done when docs/BRIEF.md exists, fits on one page, and has every section.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.