| English | 日本語 |
|---|
This repository contains prompts for product development following the "AI-Driven Development Lifecycle (AI-DLC)" process proposed by Amazon Web Services.
Prompt execution is designed with the assumption of AI agent interaction such as Amazon Q developer CLI, Claude Code, GitHub Copilot CLI, or Codex CLI.
This prompt collection consists of the following 5 phases:
- User Story Planning - Define requirements as user stories
- Unit of Work Planning - Break down user stories into implementable work units
- Domain Model Design Planning - Design domain models for business logic
- Code Implementation Planning - Implement source code from domain models
- Architecture Component Planning - Build architecture to operate the implementations
├── 01-user-story-planning.md # User story creation prompt
├── 02-unit-of-work-planning.md # Work unit division prompt
├── 03-domain-model-planning.md # Domain model design prompt
├── 04-domain-to-code-planning.md # Code implementation prompt
└── 05-architecture-component-planning.md # Architecture component creation prompt
- Select Prompt: Choose the prompt file according to the development phase
- Input to AI Assistant: Input the selected prompt to the AI assistant
- Create Work Plan: AI creates a work plan document
- Answer Questions: Answer questions in the plan document
- Approve Plan: Review and approve the work plan
- Step-by-Step Execution: AI executes step by step according to the plan
Role: Product Owner
Purpose: Create clear user stories for the target system
- 【Option】Clear context
- Describe what you want to create in
01-user-story-planning.mdand execute the prompt in the file - A work plan document for user story creation is created:
docs/inception/01-user-story/user-story-plan.md - Answer and save the questions listed in
docs/inception/01-user-story/user-story-plan.md - Instruct the AI agent to create user stories according to
docs/inception/01-user-story/user-story-plan.md - Work breakdown files related to user stories are created in
docs/inception/01-user-story/result
Role: Software Architect
Purpose: Divide user stories into work units that can be implemented in parallel
- 【Option】Clear context
- Execute the prompt in
02-unit-of-work-planning.md - A work plan document for dividing user stories is created:
docs/inception/02-unit-of-work/units-plan.md - Answer and save the questions listed in
docs/inception/02-unit-of-work/units-plan.md - Instruct the AI agent to divide user stories according to
docs/inception/02-unit-of-work/units-plan.md - User stories after work division are output to
docs/inception/02-unit-of-work/result
Role: Software Engineer
Purpose: Design domain models to implement user stories
- 【Option】Clear context
- Select the user story to be the target of domain model design (e.g.,
unit1-task-creation.md) and specify it as input in03-domain-model-planning.md - Execute the prompt in
03-domain-model-planning.md - A work plan document for domain model design from user stories is created:
docs/construction/03-domain-model/domain-model-plan.md - Answer and save the questions listed in
docs/construction/03-domain-model/domain-model-plan.md - Instruct the AI agent to divide user stories according to
docs/construction/03-domain-model/domain-model-plan.md - Domain model design documents are created in
docs/construction/03-domain-model/unitX/result
Role: Software Engineer
Purpose: Implement source code from domain model design
- 【Option】Clear context
- Select the domain model design to be implemented (e.g.,
docs/inception/03-domain-model/unitX/result/) and specify it as input in04-domain-to-code-planning.md - Execute the prompt in
04-domain-to-code-planning.md - A work plan document for implementation from domain model design is created:
docs/construction/04-domain-to-code/unitX/domain-to-code-plan.md - Answer and save the questions listed in
docs/construction/04-domain-to-code/unitX/domain-to-code-plan.md - Instruct the AI agent to divide user stories according to
docs/construction/04-domain-to-code/unitX/domain-to-code-plan.md - Implementations are created from domain model design (e.g.,
task_management)
Role: Software Architect
Purpose: Add architecture components to operate the implementations
- 【Option】Clear context
- Select the target implementation directory (e.g.,
task_management) and specify it as input in05-architecture-component-planning.md - Select the target divided user story (e.g.,
docs/inception/02-unit-of-work/result/unitX-XXX.md) and specify it as input in05-architecture-component-planning.md - Execute the prompt in
05-architecture-component-planning.md - A work plan document for architecture component creation is created:
docs/construction/05-architecture-component/architecture-component-plan.md - Answer and save the questions listed in
docs/construction/05-architecture-component/architecture-component-plan.md - Instruct the AI agent to create architecture components according to
docs/construction/05-architecture-component/architecture-component-plan.md - Architecture components are created
By executing the prompts, documents and implementations are generated in the following structure:
docs/
├── inception/ # Planning and design phase
│ ├── 01-user-story/
│ │ ├── user-story-plan.md
│ │ └── result/
│ └── 02-unit-of-work/
│ ├── units-plan.md
│ └── result/
│ ├── unit1-*.md
│ └── unit2-*.md
├── construction/ # Construction phase
│ ├── 03-domain-model/
│ │ └── unitX/
│ │ ├── domain-model-plan.md
│ │ └── result/
│ ├── 04-domain-to-code/
│ │ └── unitX/
│ │ └── domain-to-code-plan.md
│ └── 05-architecture-component/
│ └── architecture-component-plan.md
│
└── Source Code + Architecture Components