agentleFS
imtiazrayhanimtiazrayhan/agentscamp-library/skills/prd-writer/SKILL.md

prd-writer

Skill1 starsChanged 6 days agoMITView on GitHub

Turn a founder's raw idea, voice-memo transcript, or scattered bullet points into a product requirements document with a fixed section order: problem, users, jobs-to-be-done, scope, non-goals, success metrics, open questions, and risks. Use when an idea needs to be written down in a form a cofounder, contractor, or AI coding agent can build from, and it is not yet a document.

40 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use experiment-designer, user-research and 2 more.

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.
  • Dokkabei97Skill

    experiment-designer

    Designs A/B tests and product experiments with hypothesis structuring, sample size calculation, duration estimation, and result interpretation. Unified framework for both product and marketing experiments.

    Also in 3 projects that use this

  • AnthropicSkill

    user-research

    Plan, conduct, and synthesize user research. Trigger with "user research plan", "interview guide", "usability test", "survey design", "research questions", or when the user needs help with any aspect of understanding their users through research.

    Also in 3 projects that use this

  • yannickYamoSkill

    product-strategy

    Build product strategy that drives real choices. Use when defining where to play and how to win, creating strategy canvases, defining ICP/Anti-ICP, positioning products, setting pricing strategy, planning GTM motions, or structuring strategic bets. Part of the Modern Product Operating Model collection.

    Also in 3 projects that use this

  • OpenAISkill

    brainstorming

    You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

    Also in 3 projects that use this

The file

Files
SKILL.md

Most founder ideas live in a voice memo, a page of bullets, or a chat thread. None of those can be handed to a contractor, a cofounder, or an AI coding agent without the receiver filling the gaps with guesses. This skill turns whatever you have into a product requirements document (PRD) with a fixed section order, so every PRD reads the same way and the gaps are marked instead of hidden. It works on pasted text alone, with no repository, tools, or web access, so it runs identically on claude.ai, in Claude Code, and in Claude Cowork.

When to use this skill

  • You have an idea in notes form and need a document a builder can start from.
  • You are about to open an AI app builder or Claude Code and want the "what" settled before the "how".
  • A cofounder or advisor asked "what exactly are we building?" and you answered in chat.
  • You already ran the user-interview-synthesizer or competitor-teardown skill and want to fold the findings into a spec.

[!NOTE] This is a product document, not an engineering one. It says what to build, for whom, and how you will know it worked. It does not pick a stack, a database, or an architecture. For an engineering RFC grounded in an existing codebase, use the write-design-doc command after the PRD exists. In a Claude Code project, the /prd command runs this same procedure and writes the result to docs/prd.md.

Instructions

  1. Read everything the user provided before writing. Split it into statements the founder actually made and things you would have to infer. Do not ask questions up front. Write the full draft, marking every inference inline as [ASSUMPTION] and every unresolved point as [OPEN]. A founder can correct a marked guess in seconds; an unmarked one ships.
  2. Write the Problem section. One or two paragraphs: who has the problem, when it shows up, what they do about it today, and what that costs them in time, money, or risk. No solution language. If the notes jump straight to features, work backwards to the problem those features imply and mark it as an assumption.
  3. Write the Users section. Name the primary user (the person whose hands are on the product), any secondary user, and the buyer if different. One or two sentences of context each; no marketing personas.
  4. Write the Jobs-to-be-done section. Three to five statements in the form "When [situation], I want to [motivation], so I can [outcome]." Each must trace to the source material or carry an [ASSUMPTION] tag.
  5. Write the Scope section. A numbered list of what version one does, each item a behavior a user can observe ("a user can invite a teammate by email"), not a feature label ("team management"), and each testable by clicking through the product.
  6. Write the Non-goals section. What version one deliberately does not do, with a one-line reason each, including the adjacent features the founder will be tempted by. This section saves the mvp-scope-cutter skill from having to undo the PRD later.
  7. Write the Success metrics section. Two to four rows of metric, baseline, target, and time window. A baseline is a number the founder supplied or the literal text "unknown, measure first"; never invent one. Without founder numbers, targets are expressed as a direction.
  8. Write the Open questions section. Convert every [ASSUMPTION] and [OPEN] tag into a question and name who can answer it (a customer, the founder, a lawyer, an engineer).
  9. Write the Risks section. A table with columns Risk, Likelihood (low/medium/high), Impact (low/medium/high), and Mitigation. The first row is always the riskiest assumption: the single belief that, if wrong, means the product should not be built.
  10. Assemble and check. Output the PRD as Markdown with exactly these ## headings in this order: Problem, Users, Jobs-to-be-done, Scope, Non-goals, Success metrics, Open questions, Risks, in one to two pages. Re-read it once and remove any sentence that names a technology, vendor, or screen layout; those belong in the design doc.

Output

A single Markdown document with the eight fixed sections in order, a "Sources" line at the top stating what it was written from ("founder notes dated 9 Sep, one interview summary"), and inline [ASSUMPTION] tags wherever the skill filled a gap. Every tag reappears as a question in Open questions, so the founder can answer the list and re-run the skill for a clean second draft.

Example

Input: "I keep losing track of which contractors I've paid. Maybe an app where I upload invoices and it tells me what's outstanding. Small agencies would want this too."

Excerpt of the output:

## Problem
Owners of small agencies pay several contractors a month and track payments in email threads and a spreadsheet. [ASSUMPTION] They discover missed or duplicate payments only when a contractor complains.

## Success metrics
| Metric | Baseline | Target | Window |
| --- | --- | --- | --- |
| Invoices marked paid inside the product | unknown, measure first | majority of uploaded invoices | first 30 days |

## Open questions
- How many contractors does a typical target agency pay per month? (ask 5 agency owners)

The other skills in this sequence, from interview notes to a scoped v1, are described in Claude skills for founders.

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.