1---
2name: prd
3description: 'Generate high-quality Product Requirements Documents (PRDs) for software systems and AI-powered features. Includes executive summaries, user stories, technical specifications, and risk analysis.'
4license: MIT
5---
6 
7# Product Requirements Document (PRD)
8 
9## Overview
10 
11Design comprehensive, production-grade Product Requirements Documents (PRDs) that bridge the gap between business vision and technical execution. This skill works for modern software systems, ensuring that requirements are clearly defined.
12 
13## When to Use
14 
15Use this skill when:
16 
17- Starting a new product or feature development cycle
18- Translating a vague idea into a concrete technical specification
19- Defining requirements for AI-powered features
20- Stakeholders need a unified "source of truth" for project scope
21- User asks to "write a PRD", "document requirements", or "plan a feature"
22 
23---
24 
25## Operational Workflow
26 
27### Phase 1: Discovery (The Interview)
28 
29Before writing a single line of the PRD, you **MUST** interrogate the user to fill knowledge gaps. Do not assume context.
30 
31**Ask about:**
32 
33- **The Core Problem**: Why are we building this now?
34- **Success Metrics**: How do we know it worked?
35- **Constraints**: Budget, tech stack, or deadline?
36 
37### Phase 2: Analysis & Scoping
38 
39Synthesize the user's input. Identify dependencies and hidden complexities.
40 
41- Map out the **User Flow**.
42- Define **Non-Goals** to protect the timeline.
43 
44### Phase 3: Technical Drafting
45 
46Generate the document using the **Strict PRD Schema** below.
47 
48---
49 
50## PRD Quality Standards
51 
52### Requirements Quality
53 
54Use concrete, measurable criteria. Avoid "fast", "easy", or "intuitive".
55 
56```diff
57# Vague (BAD)
58- The search should be fast and return relevant results.
59- The UI must look modern and be easy to use.
60 
61# Concrete (GOOD)
62+ The search must return results within 200ms for a 10k record dataset.
63+ The search algorithm must achieve >= 85% Precision@10 in benchmark evals.
64+ The UI must follow the 'Vercel/Next.js' design system and achieve 100% Lighthouse Accessibility score.
65```
66 
67---
68 
69## Strict PRD Schema
70 
71You **MUST** follow this exact structure for the output:
72 
73### 1. Executive Summary
74 
75- **Problem Statement**: 1-2 sentences on the pain point.
76- **Proposed Solution**: 1-2 sentences on the fix.
77- **Success Criteria**: 3-5 measurable KPIs.
78 
79### 2. User Experience & Functionality
80 
81- **User Personas**: Who is this for?
82- **User Stories**: `As a [user], I want to [action] so that [benefit].`
83- **Acceptance Criteria**: Bulleted list of "Done" definitions for each story.
84- **Non-Goals**: What are we NOT building?
85 
86### 3. AI System Requirements (If Applicable)
87 
88- **Tool Requirements**: What tools and APIs are needed?
89- **Evaluation Strategy**: How to measure output quality and accuracy.
90 
91### 4. Technical Specifications
92 
93- **Architecture Overview**: Data flow and component interaction.
94- **Integration Points**: APIs, DBs, and Auth.
95- **Security & Privacy**: Data handling and compliance.
96 
97### 5. Risks & Roadmap
98 
99- **Phased Rollout**: MVP -> v1.1 -> v2.0.
100- **Technical Risks**: Latency, cost, or dependency failures.
101 
102---
103 
104## Implementation Guidelines
105 
106### DO (Always)
107 
108- **Define Testing**: For AI systems, specify how to test and validate output quality.
109- **Iterate**: Present a draft and ask for feedback on specific sections.
110 
111### DON'T (Avoid)
112 
113- **Skip Discovery**: Never write a PRD without asking at least 2 clarifying questions first.
114- **Hallucinate Constraints**: If the user didn't specify a tech stack, ask or label it as `TBD`.
115 
116---
117 
118## Example: Intelligent Search System
119 
120### 1. Executive Summary
121 
122**Problem**: Users struggle to find specific documentation snippets in massive repositories.
123**Solution**: An intelligent search system that provides direct answers with source citations.
124**Success**:
125 
126- Reduce search time by 50%.
127- Citation accuracy >= 95%.
128 
129### 2. User Stories
130 
131- **Story**: As a developer, I want to ask natural language questions so I don't have to guess keywords.
132- **AC**:
133  - Supports multi-turn clarification.
134  - Returns code blocks with "Copy" button.
135 
136### 3. AI System Architecture
137 
138- **Tools Required**: `codesearch`, `grep`, `webfetch`.
139 
140### 4. Evaluation
141 
142- **Benchmark**: Test with 50 common developer questions.
143- **Pass Rate**: 90% must match expected citations.
144