agentleFS
me-workspaceme-workspace/orchestrator-operational-agent-skills/prd-creator/SKILL.md

prd-creator

Skill1 starsChanged 45 days agoMITView on GitHub

Use this skill whenever a product requirement document is needed - "buat PRD", "tulis requirement", "spek fitur", "dokumen produk", turning an idea/chat/meeting into a formal spec, or reviewing an existing PRD for gaps. Produces a complete PRD with measurable acceptance criteria. Do NOT use for the technical design (sdlc-master) or the schedule (project-timeline-creator).

6 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

PRD Creator

A PRD's job is to make disagreement visible before code is written. It is done when an engineer can build from it without guessing and a tester can verify it without asking.

Before writing - extract these from the requester (derive from context; ask only what's truly missing)

  1. Masalah siapa, dan apa buktinya masalah itu nyata (data, keluhan, kejadian)?
  2. Apa yang terjadi kalau TIDAK dibangun? (kalau jawabannya "tidak apa-apa" - tulis itu di PRD, biar keputusan sadar)
  3. Definisi sukses dalam angka, dan kapan diukur?
  4. Batas waktu/budget yang mengikat scope?

PRD structure

# PRD: <nama fitur/produk>  (v0.x, tanggal, penulis, status: draft/review/approved)

## 1. Masalah & konteks
Siapa penggunanya, apa masalahnya, bukti masalah, biaya masalah hari ini.

## 2. Tujuan & metrik sukses
Max 3 goals; tiap goal → metrik + target + kapan diukur.
"Meningkatkan retensi" (kabur) → "D30 retention naik dari 22% ke 30% dalam 2 bulan setelah rilis" (terukur)

## 3. Non-goals (WAJIB, minimal 3)
Yang sengaja TIDAK digarap versi ini + alasan singkat. PRD tanpa non-goals = scope creep terjadwal.

## 4. User stories & alur
Per persona: "Sebagai X, saya ingin Y, supaya Z."
Alur utama langkah-demi-langkah + alur alternatif/error (pengguna salah input, koneksi putus, pembayaran gagal).

## 5. Requirements
Functional: bernomor (FR-1, FR-2, …), satu perilaku per butir, MoSCoW-tagged (Must/Should/Could).
Non-functional: performa (angka), keamanan, bahasa/lokal, device/browser, aksesibilitas, volume data.

## 6. Acceptance criteria
Per FR utama, format Given/When/Then - bisa dites, tidak ambigu:
"Given user belum bayar, When akses fitur premium, Then tampil paywall dan event `paywall_view` tercatat."

## 7. Ketergantungan & risiko
API eksternal, tim lain, keputusan legal/harga yang menggantung. Risiko + mitigasi.

## 8. Open questions
Daftar hidup; PRD tidak boleh "approved" selagi ada open question yang blocking.

## 9. Riwayat perubahan
Tanggal · apa yang berubah · kenapa - scope yang berubah diam-diam adalah pembunuh timeline nomor satu.

Quality gates before marking "review-ready"

  • Setiap requirement bisa dijawab "gimana ngetesnya?" dalam satu kalimat.
  • Tidak ada kata karet tanpa angka: "cepat", "mudah", "banyak", "user-friendly", "seamless".
  • Error/edge path tertulis, bukan hanya happy path.
  • Non-goals terisi dan sudah dikonfirmasi requester.
  • Seorang engineer yang tidak ikut diskusi bisa mengestimasi dari dokumen ini saja.

Reviewing an existing PRD

Report gaps against the structure above, urut dari yang paling berbahaya: metrik sukses tak terukur → non-goals kosong → acceptance criteria hilang → error path hilang → istilah ambigu (kutip kalimatnya, usulkan perbaikannya).

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.