agentleFS
tolomtolom/prompt-xray-skills/skills/09-prompt-to-product-spec/SKILL.md

prompt-to-product-spec

Skill1 starsChanged 5 months agoMITView on GitHub

Convert an AI prompt, assistant idea, or rough automation concept into a practical product specification with MVP scope, user stories, data sources, integrations, workflows, admin controls, and acceptance criteria.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill to turn a prompt or AI assistant idea into a product specification.

    line 10
  2. Distinctive:

    Use this skill when the user has:

    line 18
  3. Distinctive:

    If some details are missing, mark assumptions clearly.

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “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

Prompt to Product Spec Skill

Purpose

Use this skill to turn a prompt or AI assistant idea into a product specification.

The goal is to move from "the model should do this" to "the product should support this workflow".

A prompt is often only one part of an AI product. Real products also need data, UI, integrations, permissions, monitoring, and fallback behavior.

When to use

Use this skill when the user has:

  • an AI assistant idea;
  • a prompt that should become a product feature;
  • a chatbot concept;
  • an automation workflow;
  • a client request;
  • a demo idea;
  • a prototype that needs MVP definition;
  • a prompt that should be implemented by a developer or coding agent.

Required input

Collect or infer:

  • target users;
  • business goal;
  • current manual workflow;
  • model task;
  • data sources;
  • expected outputs;
  • channels or UI;
  • integrations;
  • admin needs;
  • constraints;
  • MVP deadline or budget if available.

If some details are missing, mark assumptions clearly.

Procedure

1. Extract the product goal

State the product goal in one sentence.

The product should help [user] achieve [business outcome] by using AI to [workflow].

2. Identify users and roles

Define:

  • end user;
  • admin/operator;
  • customer/client;
  • developer/maintainer;
  • AI assistant role.

3. Define current workflow

Describe how the task is done manually today.

This helps identify what should and should not be automated.

4. Define AI responsibilities

Specify what the model should do:

  • understand requests;
  • extract data;
  • classify cases;
  • generate drafts;
  • answer from sources;
  • recommend next actions;
  • prepare reports;
  • trigger workflows.

5. Define human responsibilities

Specify what humans still approve or handle:

  • final decisions;
  • sensitive replies;
  • payments;
  • legal review;
  • account changes;
  • exceptions;
  • quality control.

6. Define data sources

List:

  • user input;
  • files;
  • CRM;
  • website;
  • database;
  • email;
  • chat history;
  • knowledge base;
  • third-party APIs.

7. Define UX/channel

Specify where the feature lives:

  • web app;
  • Telegram bot;
  • website widget;
  • admin panel;
  • email workflow;
  • CRM extension;
  • internal dashboard;
  • API only.

8. Define MVP scope

Separate:

  • must-have;
  • should-have;
  • later;
  • explicitly out of scope.

9. Define acceptance criteria

For each core feature, define how to know it is done.

10. Define risks and open questions

List:

  • technical risks;
  • product risks;
  • data risks;
  • UX risks;
  • prompt/model risks;
  • unanswered questions.

Output format

Return the spec as:

# Product Specification

## One-Sentence Product Goal

The product should...

## Users and Roles

| Role | Description | Main Needs |
|---|---|---|

## Current Manual Workflow

1.
2.
3.

## Proposed AI Workflow

1.
2.
3.
4.

## AI Responsibilities

- ...

## Human Responsibilities

- ...

## Data Sources

| Source | Required | Notes |
|---|---:|---|

## User Experience

- Channel:
- Main screens/messages:
- Admin controls:

## MVP Scope

### Must-have

- ...

### Should-have

- ...

### Later

- ...

### Out of scope

- ...

## Integrations

- ...

## Acceptance Criteria

- [ ] ...
- [ ] ...
- [ ] ...

## Risks

| Risk | Impact | Mitigation |
|---|---|---|

## Open Questions

- ...

## Suggested Build Plan

1.
2.
3.

Quality checklist

A good product spec should:

  • define the business outcome;
  • separate AI and human responsibilities;
  • define MVP boundaries;
  • include data sources;
  • include acceptance criteria;
  • identify risks;
  • be actionable for implementation.

Boundaries

This skill creates a product specification. It does not estimate budget, timeline, or implementation complexity unless the required project details are available.

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.