PRD Pilot

AI workspace for product managers to turn ideas into Requirement Specs, PRDs, demo-ready HTML prototypes, and targeted iteration plans.

中文文档

What It Solves

Most AI generators can output a PRD or a mockup quickly, but the result often breaks during review:

• the PRD and demo drift apart
• key pages or flows go missing
• feedback turns into a full rewrite

PRD Pilot keeps one shared Requirement Spec across generation, validation, and iteration so product ideas can move from vague input to reviewable output with less drift.

Preview

Requirement Spec	Demo Preview	Consistency Check

PRD Draft	Targeted Iteration	Home Workspace

Quick Start

1. Start the backend

cd prd-pilot/backend
pip install -r requirements.txt
copy .env.example .env
python main.py

Example .env:

OPENAI_PROVIDER=deepseek
OPENAI_API_KEY=your_deepseek_api_key_here
OPENAI_BASE_URL=https://api.deepseek.com/v1
OPENAI_MODEL=deepseek-chat
OPENAI_MAX_TOKENS=0
APP_HOST=0.0.0.0
APP_PORT=8000

2. Start the frontend

cd prd-pilot/frontend
npm install
npm run dev

3. Open the app

• Frontend: http://localhost:5173
• Backend health: http://localhost:8000/api/health

Core Features

Shared Requirement Spec

PRD Pilot first structures user input into a single internal spec:

• product_name
• product_type
• target_users
• user_pain_points
• core_scenarios
• key_features
• primary_pages
• user_flow
• style_preference
• constraints
• success_criteria

This spec becomes the shared source of truth for every later step.

PRD, Demo, and Prototype Outline

• PRD: Chinese Markdown draft for requirement review
• Demo: single-file HTML prototype for direct preview and download
• Prototype Outline: page structure, flow, and validation goals

Consistency Check

Built-in checks cover:

• page coverage
• feature coverage
• flow connectivity
• naming consistency
• prototype alignment
• scenario coverage

Targeted Iteration

Instead of regenerating everything, PRD Pilot supports scoped updates such as:

• add page
• modify user
• remove feature
• adjust layout
• change style
• improve data density
• simplify PRD
• clarify flow

Each iteration returns a short change summary.

In-Browser Model Configuration

The UI supports page-level model configuration for OpenAI-compatible APIs:

• provider
• model name
• API key
• base URL
• max tokens (optional; blank means auto)

Built-in presets:

• DeepSeek
• OpenAI
• OpenRouter
• Zhipu / GLM
• SiliconFlow
• Moonshot
• Groq
• DashScope / Qwen
• Ollama (Local)
• Custom OpenAI Compatible

Workflow

flowchart LR
    A[Input Idea] --> B[Structure Requirement Spec]
    B --> C[Generate PRD]
    B --> D[Generate Demo HTML]
    B --> E[Generate Prototype Outline]
    C --> F[Consistency Check]
    D --> F
    E --> F
    F --> G[Targeted Iteration]
    G --> B

Loading

Who It's For

• student product managers who need to prepare requirement reviews quickly
• indie developers who need reviewable specs and demo-ready prototypes
• small teams without dedicated design or frontend prototyping support

API

• GET /api/model-options
• POST /api/test-model-config
• POST /api/structure-requirement
• POST /api/generate-prd
• POST /api/generate-demo
• POST /api/check-consistency
• POST /api/iterate-prd
• POST /api/iterate-demo
• GET /api/health
• GET /api/test-llm

Tech Stack

Frontend

• Vue 3
• Vite
• Element Plus
• Tailwind CSS
• MarkdownIt
• VueUse

Backend

• FastAPI
• OpenAI-compatible API client
• Pydantic
• Python Dotenv

Project Structure

.
├─ prd-pilot/
│  ├─ backend/
│  │  ├─ main.py
│  │  ├─ requirements.txt
│  │  └─ services/
│  │     └─ llm_service.py
│  └─ frontend/
│     ├─ src/
│     │  └─ App.vue
│     ├─ package.json
│     └─ vite.config.js
├─ docs/
│  ├─ README.zh-CN.md
│  └─ screenshots/
├─ README.md
└─ LICENSE

Current Scope

• prototype output is HTML Demo + Prototype Outline
• no image-based prototype generation yet
• no persistent version rollback yet
• consistency check v1 is rule-based, not AI-score-driven

License

MIT