AI · Prompt
PRD Prompt for AI Coding Agents
Prompt for generating granular, agent-ready product requirements documents in Markdown
A structured prompt designed to turn an app concept into an actionable, highly detailed Product Requirements Document formatted in Markdown. It specifies a modern tech stack (Next.js, Drizzle, Supabase, Shadcn) and forces the AI to output granular 10-to-15-minute implementation subtasks for coding agents like Cursor or v0.
The prompt
**Task:** Create a comprehensive Product Requirements Document (PRD) using the guidelines below. This document should be designed for guiding AI coding agents and formatted in **Markdown syntax**.
### Context and Requirements
**Tech Stack:**
- **TypeScript**, **Next.js 15**, **Drizzle ORM** with **Supabase**, **React Query**.
- **UI Design**: **ShadCN UI**, **Tailwind CSS**, **Lucide Icon**.
- **Additional Tools**: Use **Next.js server actions** for database interactions. If AI features are required, leverage **OpenAI's API** and **Vercel AI SDK**.
### Structure of the PRD
1. **Features**:
- Begin by categorizing features into:
- **Key Features**: Absolutely essential for the MVP.
- **Nice-to-Have Features**: Enhancements that improve usability but are not critical.
- Example:
- Key Feature: Chat functionality.
- Nice-to-Have: Search past chat history.
2. **Routes and UI Components**:
- Brainstorm all required routes.
- For each route, include:
- **Core Purpose**: The primary function of the route.
- **Key UI Components/Inputs**: List components and their purposes.
- Example:
- /dashboard
- **Core Purpose**: Display user metrics for admins.
- **Components**: Metric cards, graph for interactions over time.
3. **Third-Party Tools**:
- Identify **APIs or libraries** needed for non-native features.
- Clearly describe their **purpose** and **integration strategy**.
4. **Database Schema**:
- Provide an **initial schema** for database tables, considering the tech stack's requirements.
5. **State Management**:
- Outline key **TypeScript types** or **interfaces** for objects and props.
6. **Component Architecture**:
- Map out components with clear **inputs, outputs, and data sources**.
- Example:
- prompt-input:
- **Purpose**: Handle shortcuts fetched from the database and control user input.
- **Integration**: Used in multimodal input components.
7. **Implementation Plan**:
- CRITICAL REQUIREMENTS FOR IMPLEMENTATION PLAN:
- FOCUS EXCLUSIVELY ON MUST-HAVE FEATURES FOR THE MVP
- Create THE MOST DETAILED, GRANULAR implementation plan possible
- Each task must have 5+ HIGHLY SPECIFIC sub-steps
- Include EXACT technical implementation details
- Break down EVERY single step into actionable 10-15 minute tasks
- Write as if instructing a junior developer who needs EXPLICIT, STEP-BY-STEP guidance
- Break the implementation into **tasks** and **steps**:
- Label tasks numerically (e.g., ### 1. Create Dashboard).
- Break down steps under each task with EXTREMELY GRANULAR subdivisions
- Make steps as specific as possible. FOR EXAMPLE:
- INSTEAD OF: "create a route and page"
- DO: "create a Next.js route in `app/upload/page.tsx` that includes a file input component using ShadCN UI. Add client-side validation to check for CSV file type and maximum file size of 10MB. Implement drag-and-drop functionality with clear error messaging for invalid files."
- Prioritize tasks to minimize time to interaction (e.g., use placeholders or dummy data initially).
8. **Documentation Section**:
- Leave a section for adding details on any relevant APIs, libraries, or tools used, that are not already part of our tech stack.
### Additional Guidelines
- **Ask Clarifying Questions**: Begin by addressing any uncertainties or gaps in the provided information.
- **Expand Features or Topics**: Feel free to suggest additional features or improvements, but **do not omit any provided requirements**.
- **Precise Technical Details**: Ensure all technical descriptions are thorough and actionable.
### IMPLEMENTATION PLAN CRITICAL INSTRUCTIONS
- GRANULARITY IS THE ABSOLUTE KEY
- Every single task must be broken into the smallest possible actionable steps
- Describe EXACTLY how each step will be implemented
- NO BROAD OR VAGUE DESCRIPTIONS ALLOWED
What it's for
- Generate a structured MVP spec before starting a build in Cursor or Windsurf
- Define routes, UI component hierarchies, and database schemas for v0 or Bolt.new
- Break down full-stack development tasks into granular sub-steps for AI coding agents
Key points
- Preconfigured modern web stack focusing on Next.js 15, Drizzle ORM, Supabase, and Shadcn UI
- Feature prioritization splitting requirements into core MVP vs nice-to-have items
- Detailed route mapping paired with core component and input breakdowns
- Initial database schema and TypeScript interface planning
- Component architecture mapping with inputs, outputs, and data sources
- Enforced ultra-granular implementation plan with 5+ explicit sub-steps per task
Pricing
Free
Details
- Type
- Prompt
- Category
- AI
- Topic
- AI Prompts
- Open source
- No
- Platforms
- Web
- For
- Developers and technical founders building apps with AI coding assistants
- Tags
- prompt, cursor, prd, ai coding, nextjs, spec
Why it was saved
- @kai_rypto25: Reusable prompt for turning an app idea into a granular, agent-ready spec for tools like V0 or Cursor.
Related
Curated library of specialized AI agent personas and workflows for Claude Code
System prompt to convert rough software ideas into structured, agent-ready PRDs
A cheat sheet of 15 business and marketing prompts for ChatGPT.
Three-step transactional analysis prompt sequence for forecasting your future life trajectory
A collection of reusable ChatGPT prompts designed for studying and mastering topics
A system prompt that removes conversational fluff and engagement-seeking behaviors from LLMs
Updated