Back to AI library

Skill

Product requirements document (PRD)

Draft product requirements from your existing context with this AI skill. Use it early in planning to create a quick spec or PRD with the level of detail your idea supports.

Open this AI skill

About this AI skill

The Product requirements document (PRD) skill is pre-built for Elle (the AI assistant in Aha! software) and reviews your notes, linked product records, and reference material before drafting requirements. It produces a quick spec, minimum viable PRD, full PRD, or focused brief based on your request and the available evidence, preserving your template when you provide one. The draft defines the user problem and proposed scope, labels assumptions and open questions, and identifies what needs validation before you proceed.

The full Product requirements document (PRD) skill, as Elle uses it:

# Product requirements document (PRD)

Use when you want to draft product requirements in the lightest useful format, working from available context before asking questions.

## Definitions

Use these conventions unless the user asks for a different format.

- **Status.** Not started, On track, At risk.
- **User story format.** As a [type of user], I want to [action], so that [benefit].
- **Acceptance criteria format.** Given [how things begin], when [action taken], then [outcome].
- **Table format.** Two columns, no header row, row name in the first column and content in the second.
- **Evidence labels.** Known from context, Assumption, Open question, Needs validation.
- **Output modes.** Quick spec, minimum viable PRD, full PRD, or a focused release, epic, or feature brief.

## Workflow

1. Check context first. Read any provided note, prompt, linked record, reference document, template, research, or feedback. A note holding high-level product context, such as a vision statement, is the strongest input and lets you skip most clarifying questions. Prefer this source order: user-provided material, then linked product context such as goals and initiatives and ideas, then reference templates, then clearly labeled assumptions. If a template is provided, preserve its structure and terminology.
2. Choose the lightest output mode that fits the request and the maturity of the idea
3. Ask at most 1 to 3 high-value questions, and only when the missing information would materially change the draft. Prioritize target user, core problem, desired outcome, available evidence, and required level of detail.
4. Refine the problem before drafting. Produce a short synthesis of product area, target user, user challenge, business outcome, evidence, assumptions, and open questions. If the idea is still too vague, ask the next best question instead of generating a bloated draft.
5. Draft an evidence-backed document. Include only the sections the chosen mode calls for.
6. Expand only when warranted. Generate the delivery hierarchy the evidence supports and no more. Mark speculative breakdown items as draft recommendations.
7. Review before finalizing. Check for missing goals or success metrics, unclear user or problem definition, weak evidence, missing non-goals, hidden dependencies, and decisions that need design, engineering, or customer validation. End with a short **Needs validation** section and suggested next steps.

## Rules

- Be direct and practical. Favor a smaller, sharper draft over a comprehensive but speculative one.
- Label every missing or inferred detail with **Assumption** , **Open question** , or **Needs validation**
- Create a markdown document artifact for substantial output
- Use short sections and bullets for a quick spec. Use tables where they improve scannability.
- When the user asks for a PRD based on a template, preserve the template shape rather than inventing a new one
- Include metrics and analytics considerations when the request implies measurable or customer-facing behavior
- Headers use sentence case. No em dashes, en dashes, or curly quotes.

## Output

**Quick spec:** Overview, problem, target user, desired outcome, proposed approach, risks, open questions

**Minimum viable PRD:** Overview, problem statement, goals and success metrics, target users, scope, non-goals, core requirements, dependencies and risks, open questions, needs validation

**Full PRD:** Overview, problem statement, goals and success metrics, background and evidence, users or personas, use cases or user stories, scope and non-goals, functional requirements, non-functional requirements when relevant, dependencies and constraints and risks, UX considerations, launch considerations, open questions, needs validation

**Release, epic, and feature artifacts:** Tie each item to the user problem and business goal it supports. Add user stories and acceptance criteria only when the available evidence grounds them.

## Quality checks

- Every inferred detail carries an evidence label
- Goals have measurable success metrics
- Scope is bounded by explicit non-goals
- The document ends with needs validation and next steps

Try this AI skill now

Open the Product requirements document (PRD) skill in Aha! Roadmaps to turn your product context into a requirements draft.

Open this AI skill

About Elle

Elle , the AI assistant in Aha! software, helps product teams accelerate product work by creating content, analyzing records, and updating plans using information in their account. The AI prompt library offers pre-built skills and tasks for Elle that are tuned for product development work.