| name | create-prd |
| description | Create a Product Requirements Document from conversation. Use when starting a new project or defining product requirements. |
| disable-model-invocation | true |
| allowed-tools | Read, Write, Glob, Grep |
| argument-hint | ["project-name"] |
Create PRD: Generate Product Requirements Document
Overview
Generate a comprehensive Product Requirements Document (PRD) based on conversation context. This PRD will serve as the single source of truth for project-specific requirements.
Important: The PRD is the ONLY project-specific file. All other Agent Kit files are reusable.
Output File
Write the PRD to: .claude/PRD.md
If a PRD already exists, ask before overwriting.
Before Starting
1. Gather Project Context
Ask the user for (if not already discussed):
- Project Name: What is this project called?
- Client/Domain: Who is this for? What industry?
- Core Problem: What problem are we solving?
- Target Users: Who will use this?
- MVP Scope: What's the minimum viable product?
2. Reference Agent Kit Standards
Read these files to ensure PRD aligns with our standards:
CLAUDE.md - Tech stack and code standards
.claude/reference/architecture.md - Architecture patterns
.claude/reference/design-system.md - UI/UX standards
.claude/reference/mastra-best-practices.md - AI agent patterns
PRD Structure
Create a well-structured PRD with the following sections:
# [Project Name] - Product Requirements Document
> **Note:** This PRD contains project-specific requirements.
> All technical standards are inherited from Agent Kit (`CLAUDE.md`).
**Version:** 1.0
**Status:** Draft | In Progress | Production-Ready
**Last Updated:** [Date]
---
## Table of Contents
1. [Executive Summary](#executive-summary)
2. [Vision & Principles](#vision--principles)
3. [Problem & Solution](#problem--solution)
4. [Target Users](#target-users)
5. [MVP Scope](#mvp-scope)
6. [User Stories](#user-stories)
7. [Feature Specification](#feature-specification)
8. [Domain Model](#domain-model)
9. [AI Agent Specification](#ai-agent-specification)
10. [API Specification](#api-specification)
11. [Implementation Phases](#implementation-phases)
12. [Success Criteria](#success-criteria)
13. [Risks & Mitigations](#risks--mitigations)
---
## Executive Summary
### Vision
[2-3 sentences describing the product vision]
### The Difference
| Traditional Approach | Our Approach |
|---------------------|--------------|
| [Old way] | [New way with AI] |
### What We're Building
[Concise description of the product]
### MVP Goals
**Goal:** [One sentence MVP goal]
**Success measured by:**
- [Metric 1]
- [Metric 2]
- [Metric 3]
---
## Vision & Principles
### Core Principles
1. **[Principle 1]**
- [Description]
2. **[Principle 2]**
- [Description]
3. **[Principle 3]**
- [Description]
---
## Problem & Solution
### The Problem
[Detailed problem description]
### Our Solution
[How we solve it]
### Key Differentiators
- [Differentiator 1]
- [Differentiator 2]
---
## Target Users
### Primary Persona
**[Persona Name]**
- Role: [Job title/role]
- Tech Comfort: [Low/Medium/High]
- Key Needs:
- [Need 1]
- [Need 2]
### Secondary Personas
[If applicable]
---
## MVP Scope
### In Scope
**Core Functionality:**
- [Feature 1]
- [Feature 2]
**Technical:**
- [Tech requirement 1]
- [Tech requirement 2]
### Out of Scope
- [Deferred feature 1]
- [Deferred feature 2]
---
## User Stories
### Primary Stories
**US-1: [Story Title]**
> As a [user type], I want to [action], so that [benefit].
**Acceptance Criteria:**
- [ ] [Criterion 1]
- [ ] [Criterion 2]
**US-2: [Story Title]**
> As a [user type], I want to [action], so that [benefit].
[Continue for 5-8 primary stories]
---
## Feature Specification
### Feature 1: [Feature Name]
**Purpose:** [Why this feature exists]
**Functionality:**
- [Function 1]
- [Function 2]
**UI/UX:**
- [UI element 1]
- [UI element 2]
[Continue for each feature]
---
## Domain Model
### Key Entities
| Entity | Description | Key Fields |
|--------|-------------|------------|
| [Entity 1] | [Description] | id, name, ... |
| [Entity 2] | [Description] | id, name, ... |
### Entity Relationships
[Entity A] 1--* [Entity B]
[Entity B] *--1 [Entity C]
### Enums/Categories
| Enum | Values | Description |
|------|--------|-------------|
| [Category] | value1, value2, value3 | [Purpose] |
---
## AI Agent Specification
### Agent: [Agent Name]
**Purpose:** [What the agent does]
**Personality:**
- [Trait 1]
- [Trait 2]
**Capabilities:**
1. [Capability 1]
2. [Capability 2]
**Tools:**
| Tool | Purpose | Input | Output |
|------|---------|-------|--------|
| [Tool 1] | [Purpose] | [Input] | [Output] |
### Agent Instructions (System Prompt Summary)
[Key points from system prompt]
---
## API Specification
### Endpoints
| Method | Endpoint | Purpose |
|--------|----------|---------|
| GET | /api/[resource] | [Purpose] |
| POST | /api/[resource] | [Purpose] |
### Example: [Endpoint Name]
**Request:**
```json
{
"field": "value"
}
Response:
{
"id": "uuid",
"field": "value"
}
Implementation Phases
Phase 1: Foundation
Goal: [Phase goal]
Deliverables:
- [Deliverable 1]
- [Deliverable 2]
Validation:
- [How to verify completion]
Phase 2: Core Features
Goal: [Phase goal]
Deliverables:
- [Deliverable 1]
- [Deliverable 2]
Phase 3: Polish & Launch
Goal: [Phase goal]
Deliverables:
- [Deliverable 1]
- [Deliverable 2]
Success Criteria
MVP Success Definition
Functional Requirements:
- [Requirement 1]
- [Requirement 2]
Quality Indicators:
- [Indicator 1]
- [Indicator 2]
Metrics
| Metric | Target | Measurement |
|---|
| [Metric 1] | [Target] | [How measured] |
Risks & Mitigations
| Risk | Likelihood | Impact | Mitigation |
|---|
| [Risk 1] | Medium | High | [Mitigation strategy] |
| [Risk 2] | Low | Medium | [Mitigation strategy] |
Technical References
Note: Technical implementation details are in Agent Kit reference files.
| Topic | Reference |
|---|
| Tech Stack | CLAUDE.md |
| Architecture | .claude/reference/architecture.md |
| Design System | .claude/reference/design-system.md |
| AI Agents | .claude/reference/mastra-best-practices.md |
| Deployment | .claude/reference/deployment-best-practices.md |
Version History:
---
## Instructions
### 1. Extract Requirements
- Review the entire conversation history
- Identify explicit requirements and implicit needs
- Note technical constraints and preferences
- Capture user goals and success criteria
### 2. Synthesize Information
- Organize requirements into appropriate sections
- Fill in reasonable assumptions where details are missing
- Maintain consistency across sections
- Ensure alignment with Agent Kit standards
### 3. Write the PRD
- Use clear, professional language
- Include concrete examples
- Use markdown formatting (headings, lists, code blocks, checkboxes)
- Keep Executive Summary concise but comprehensive
### 4. Quality Checks
- All required sections present
- User stories have clear benefits
- MVP scope is realistic
- Aligns with Agent Kit tech stack (CLAUDE.md)
- Implementation phases are actionable
- Success criteria are measurable
## After Creating PRD
### 1. Update PROJECT-STATUS.md
```markdown
## Project Info
- **Project:** [Name]
- **PRD:** Created [date]
- **Phase:** Planning
2. Output Confirmation
Provide:
- Confirm the PRD was written to
.claude/PRD.md
- Brief summary of contents
- Assumptions made (if any)
- Suggested next steps
3. Next Steps
Recommend:
1. Review the PRD and refine if needed
2. Run /prime to load project context
3. Start planning with /plan-feature [first-feature]
Notes
- If critical information is missing, ask clarifying questions BEFORE generating
- The PRD should be project-specific; technical standards come from CLAUDE.md
- For AI agent projects, emphasize the Agent Specification section
- For CRUD apps, emphasize Feature Specification and API sections